FOR THE PEOPLE FOR EDVCATION FOR SCIENCE LIBRARY OF THE AMERICAN MUSEUM OF NATURAL HISTORY Charles Otis Whitman 1901 Age 60 years ORTHOGENETIC EVOLUTION IN PIGEONS POSTHUMOUS WORKS OF CHARLES OTIS WHITMAN PROFESSOR OF ZOOLOGY IN THE UNIVERSITY OF CHICAGO, I892-IQIO; DIRECTOR OF MARINE P C BIOLOGICAL LABORATORY AT WOODS HOLE, [888-I908 EDITED BV OSCAR RIDDLE RESEARCH ASSOCIATE OF THE CARNEGIE INSTITUTION OF WASHINGTON VOLUME I PUBLISHED BY THE CARNEGIE INSTITUTION OF WASHINGTON Washington, 1919 CARNEGIE INSTITUTION OF WASHINGTON Publication No. 257, Volume I Paper No. 28 or the Station for Experimental Evolution at Cold Spring Harbor, New York PRESS OF J. B. LIPPINXOTT COMPANY PHILADELPHIA PREFACE BY THE EDITOR. At the mature age of 50 years Professor Charles Otis Whitman began the work now presented in these volumes, and the long task was still incomplete when, 18 years later, on December 6, 1910, a brief illness terminated the work and life of this investigator, to whom biology and many American biologists especially are so much indebted. After six years, the efforts of some of his friends to make the results of his long labors available to others are nearly concluded. This task has been necessarily shared by a number of persons and organizations, and the demands upon some of them have been so considerable and unusual that the nature of their endeavors is here most gratefully acknowledged. The continued maintenance of the large collection of pigeons upon which this series of studies was based was essential to a presentation of anything more than a fragment of the results; it was necessary to learn the term of life and the sex of all the surviving individuals; and further breeding or testing of certain "mutants" among them was scarcely less urgent. This formed one of a number of heavy exactions upon those who have felt the necessity of presenting the largest possible portion of the author's results. The whole of these and other obligations, during the first year, was shared by Mrs. C. 0. Whitman and by the editor, who was generously assisted by unasked grants from the A. A. Sprague Institute of Chicago and from the Laboratory of Experimental Therapeutics of the University of Chicago. During the five following years the Carnegie Institution of Washington has fully maintained the work — during two years at Chicago and for the succeeding three years at the Station for Experimental Evolution at Cold Spring Harbor, New York; that Institution has also assumed the entire expenditures incident to the publication of these volumes. For all this generous assistance the editor expresses enduring obligations. Not more than one-fifth of the matter herewith presented was found to approach a stage of readiness for publication; it has required some years to bring the results to their present form. Even now we have been unable to utilize all of the records and data, and many of the carefully prepared illustrations are also omitted. If our progress in the preparation of these materials has seemed lingering and slow, it should be said that much time has been given to all those materials which proved impracticable for use as well as to those which are now incorporated in the work. It was, however, in some cases found obligatory to become familiar with a subject as to which only records and raw data were avail- able. Again, other and nameless difficulties have sometimes attended our efforts. In the task of preparing the various materials for publication the editor found himself wholly unprepared to deal with the manuscripts on Behavior in Pigeons. Professor Harvey A. Carr has most generously carried that work to completion and these studies are grouped to form Volume III. The special materials on voice were iy PREFACE BY THE EDITOR. assigned to Professor Wallace Craig, who was able to arrange for publication only a small part of them; these also will be found in Volume III. The dominant feature of Professor Whitman's prolonged study of inheritance and evolution lies in its intensive and diversified attack upon the nature of a specific character. As early as 1897 he wrote: It is to a comparative and experimental analysis of specific characters that we must look for a knowledge of the phenomena of heredity and variation. As Whitman was a contemporary of deVries, who likewise avowed that the study of the origin of species is to be accomplished through a study of specific characters, it is truly remarkable that the results of the two investigators should be so strongly opposed. The reasons for these divergent conclusions may be demon- strated at some time in the future, but something can now be said as to differences in method and material, which certainly have accompanied the two decisions which now seem to stand in strong disagreement. It is clear, for example: (1) That deVries worked with much larger numbers of individuals than did Whitman. A corollary of this is — though it may not at first be entirely evident to all — that Whitman was able to and did more closely study and observe the specific characters with which he worked, and for longer periods of time, in the same indi- vidual. The one study was expansive, the other intensive. (2) The purity of the greater part of Whitman's breeding material is beyond question. Can the same be said for de Vries's material of chief reliance? (3) The phylogenetic relationships of Whitman's most studied species were known and the direction taken by evolution in the past had been ascertained. Does the want of this orientation in de Vries's material at all qualify his results? (4) The observations and resultant theory of de Vries nowhere touch or are con- cerned with recapitulation. Whitman finds this to be the central fact of heredity and organic development, and the specific characters most carefully studied by him reveal it abundantly. Does this difference lend weight to the conclusions of the one or of the other? Professor Whitman's devotion to the task of mastering a specific character was not limited by the conventions of any particular line of study; it heeded neither time, personal sacrifice, nor the knotty and thorny interpolations which the ensemble of life-processes is continually throwing upon the path of the biologist when he would isolate and examine a particular vital process; and, surely, these latter difficulties especially beset all studies on the origin and establishment of new forms of organisms. But Whitman was ready and eager to attend to each and every intercalated phenomenon, from whatever foreign or extrinsic source, if its analysis and meaning might lead to a better, surer, or closer measure of realities in his own main sphere of study. It thus happens that along the pathway which he has blazed into the central problems of evolution are to be found other landmarks of discovery — some mere sign-posts and some wide and well-done surveys of regions which lead well into the territory of such other and diverse subjects, as instinct, fertility, animal behavior, correlative variation, and the nature of sex. PREFACE BY THE EDITOR. V Having selected color-pattern in pigeons as supplying a relatively small group of specific characters easily accessible to study, he first set about determining which patterns are the more primitive and which the higher and more recent, each bit of evidence being retested by search for the convergent testimony of various kinds of evidence. A general survey was made of the color-patterns of the wild species of pigeons, a group that includes nearly 600 species, besides nearly 200 domestic races that have been derived from one or a few wild species. Large numbers of genera and species from all parts of the world were brought to the breeding-pens of his yard. There with much patience the patterns were studied from the living birds; male and female were compared; the sequence of pattern in the plumages from young to old was accurately observed; thoughtful experiments were devised to bridge the gap between the molts and thus displace apparent discontinuities with evident continuities; the behavior of many characters in hybridization was exten- sively studied; and the primitive pattern of several diverse orders of birds was ascertained. The direction of the evolution of pattern as it was indicated by all these studies was, moreover, further tested by evidence of an entirely different sort. Such char- acters as voice, behavior, and fertility were separately subjected to similar appro- priate, comparative, and breeding tests to learn whether the resulting data would parallel each other and whether all would parallel the data furnished by the exten- sive study of the color-pattern. Only when Dr. Whitman had accumulated a vast amount of consistent and convergent testimony as to where the various genera stand in the phylogenetic series did he permit himself to feel that he was reading aright the history of the specific characters of the pattern. In consequence, Whitman's work presents a large body of searchingly self- critical conclusions; and these conclusions unquestionably lead far into constructive evolutionary theory. For his material, he was convinced that he had demonstrated the reality and regnancy of definitely directed variation — i.e., of orthogenesis — as the method of evolution. He has accumulated the most weighty evidences for continuity as against discontinuity in the phenomena of variation, inheritance, and evolution. He has thrown new and extraordinary light on the nature and meaning of "mutants" — such "mutants," at any rate, as occur among pigeons; and he has made a brilliant and comprehensive analysis of the phylogeny of pigeons. Chapters I, II, III, IV, and X of the present volume were practically finished by Dr. Whitman, and except for the addition of references to the illustrations and the occasional insertion of a page or a few pages of supplementary matter — often written later by him on the same subject — these remain practically unchanged. The illustrations of Chapter X were descriptively united to the text by Dr. Whitman. Nearly all of the other illustrations have been placed by the editor at those points in the text which to him seemed necessary or appropriate. Chapters V to IX inclusive required a complete rearrangement by the editor. The sources of this material were quite scattered, as is indicated in connection with each chapter, and some topics were not sufficiently treated. In other cases the subjects were studied in the earlier years and it was found that later and more searching study had VI PREFACE BY THE EDITOR. thrown other or additional light on these topics. Several complete addresses by Dr. Whitman were found to repeat statements elsewhere given ; these have been either entirely omitted or certain paragraphs have been transferred to a position in one or another of the present chapters where the same topic is discussed. In some instances the editor has added statements of his own, but it is thought that the reader will never be in doubt as to the source of a statement. In all those chapters in which the present arrangement is wholly that of the editor, and where he has drawn material from manuscripts or folders variously designated by the author, references to those original folders are made at various points within the chapter or at its close. All of these original folders will be permanently filed with the Carnegie Institution of Washington, either in Washington or at the Station for Experimental Evolution at Cold Spring Harbor, where they will at all times be easily accessible to all interested parties. The reader will find that the literature of some of the subjects considered here is not adequately treated. It will be easily granted, however, that since the author did not arrange these parts of the manuscript for publication, it would be unwise or impossible for another to go far in an attempt to do so. To the two artists, Mr. K. Hayashi and Mr. Kenji Toda, whose skilful and numerous productions assist so much in the presentation of these studies, the author would, I am sure, have recorded his warmest gratitude. The editor readily acknowledges that the imperfections of form of the work as it now appears are assignable to him, and not to the author; but if what he has done lacks the finish and completeness of all that which the author had made ready for publication, it is hoped that a full measure of indulgence may be granted for imper- fect results in the rather difficult endeavor to make available material that other- wise must have been wholly lost. But even while deploring the imperfections of his work, he avows it — under the circumstances — the great pleasure of his life to have been able to assist in presenting the message of his colleague and master to those for whom the labors of many years had pledged it. Oscar Riddle. Station for Experimental Evolution, Cold Spring Harbor, December 1916. TABLE OF CONTENTS. Page Preface by the Editor iii Chapter I. Introductory 3 Chapter II. The Problem of the Origin of Species 9 Chapter III. The Problem of the Origin of Species (continued) 37 Chapter IV. The Origin and Relationship of Rock-Pigeons as Revealed in their Color-Pattern. ... 49 Chapter V. The Turtle-dove Pattern in the Piiylogeny of Pigeons 64 General statement 64 The turtle-dove pattern in the 1 'nisi cliche 66 The Turturinse 66 The Zenaidina- 75 The Peristerinie 77 The Geopeliinx 84 The Phabinae 95 The Geotrygoninse ' 102 Chapter VI. The Turtle-dove Pattern in the Phylogeny of Pigeons (continued) 104 The turtle-dove pattern in the Treronidse 104 The Treronina; 104 The Ptilopodina; 104 The turtle-dove pattern in the Gouridse 105 The turtle-dove pattern in the Columbidae 105 Feral Columba 105 Domestic pigeons 112 Color and pattern in domestic pigeons, according to Priitz 113 Black crescentic tips in toy pigeons 115 Chapter VII. The Turtle-Dove Pattern in Other Orders of Birds 117 Historical and introductory 117 Theories of color and pat tern 11' Sexual dimorphism in color 120 The turtle-dove pattern and its modification in various forms 121 On the evolution of the ocellus of the Argus pheasant and tin' peacock 131 Further examples of the turtle-dove pattern and its derivatives in birds .134 Dark centers 134 Dark centers and transverse bars 134 Transverse bars '35 Black crescentic tips and transverse bars 136 Concentric crescentic bars 138 Transverse bars in common fowls 138 Chapter VIII. Funis and Fundamental Bars as Plumage Characters .... 140 Fluted and frilled wing-feathers 1-10 Common pigeon and guinea-pigeon 141 Flutes pass into frills 141 The association of frills with other evidences nf weakness 142 The breast-crease and frill "3 The breast-crease in hybrids from the common pigeon and the Japanese turtle 143 The breast -crease and frill in other hybrids and in feral species 145 The breast-crease and frill in common pigeons 14< < In frilled or frizzled races of birds The pigeon 1 Fowls I 50 Geese ' ^ Bullfinch • l51 Frilled feathers as a character of gradual development rather than saltative 151 Fundamental bars 15- Chapter IX. The Mutation Theory and Mutations 15" Introductory '" The guinea-pigeon mutation *°" The Zenaida mutation "" On the color of the young of Zenaida and Zenaidura 168 On the juvenal plumage colors of a " mutant " and normal Zenaida 171 General considerations Chapter X. The Problem of Organic Development — Facts and Theories 177 The apical mark as an example of recapitulation 182 vii ILLUSTRATIONS. PLATES. Frontispiece, Charles Otis Whitman in 1901, al the age of 60 years. facing page 1. A. Adult mule two-barred rock-pigeon, Columba lit ia. B, Vdult male chequered wild rock-pigeon, ( '. affmis. 12 2. Adult C. linn, uniform gray third bar H 3. A, Fully chequered wing of domestic pigeon about 6 weeks old. B, Chequers more strongly reduced in the anterior part of the wing of an adult homer 10 4. A, Male homer, No. 111. B, female homer, Xo. K2 -'(I 5. A, Female homer, No. 3. B, Male homer, No. 7. C, Female homer. No. 8 22 6. A, Left wing of bronze-winged pigeon, Phaps chalcoptera. B, Adult male bronze-wing, Phaps elegans 26 7. A, Wing of juvenal mourning-dove, Z\ naidura carolinensis. It. Wing of juvenal female bronze-wing, Phaps chalcoptera 28 8. A, Wing of adult male' crested pigeon, Ocyphaps lophotes. 13, Wing of juvenal female crested pigeon, 0. Inphotes 30 9. Adult female stock-dove, Columba anas 32 10. Left wing of adult stock-dove, C. anas, No. 2 32 11. Left wing and bars of both wings of adult stock-dove, I '. mnas, Xo. 3 32 12. A, Wing-bars of adult stock-dove, C. anas, No. 5. B, Wing-bars of adult stock-dove, C. anas, No. 11 34 13. Rudimentary fourth wing-bar in adult female stock-dove, No. 24 34 14. Juvenal female crested pigeon, Ocyphaps lophott s, age 32 days 42 15. Adult female crested pigeon, 0. lophotes 42 16. Adult male crested pigeon, O. lophotes 44 17. The longer left wing-coverts of adolescent (second plumage) male, adult male, and adult female crested pigeons 46 18. Adult spotted pigeon, Columba maculosa 60 19. A, Juvenal, 6 weeks old, Damar ring-dove, St. damarensis. B, Juvenal blond ring, St. risoria 68 20. The original feather color-pattern, as shown in a female Japanese turtle-dove, Turtur orientalis 70 21. Feather pattern in the European turtle-dove, Turtur turtur 70 22. 1, Left wing of an adult European turtle-dove, T. turtur. 2, Third longer covert of juvenal T. oru nialis. 3, Third longer covert of juvenal T. turtur. 4, Third longer covert of juvenal Si damarensis. 5, Third longer covert of juvenal Slig. senegalensis 72 23. A, Adult Chinese turtle-dove, Spihpt lia chiru n&is. IS, Fifth feather, third row, of the neck-mark of Chinese turtle-dove. C, Wing of juvenal mourning-dove and ring-dove hybrid 72 21. A and B, Scapulars and tertials from adult female Zenaidura carolinensis ill I. C and D, Scapulars and tertials from adult male Zenaida vinaceo-rufa (11). E, Third tertial from a hybrid (ZZ-A1) 7ti 25. Selected feathers from juvenal mourning-dove, Zenaidura carolim nsis 70 26. A, Neck-mark of adult male white-winged pigeon, Melopi lia It ucopti ra. B, Neck-mark of adult male mourn- ing-dove, Zenaidura carolim nsis 7t> 27. A, Fifteen feathers from adult male Chamapelia passerina. B, Five feathers from adult female Ch. passerina 78 28. A, Left wing of juvenal female Ch. passerina. 1*>. Feathers corresponding to figures of the male shown in plate 27. C, Selected feathers from a juvenal Ch. passi rina 80 29. A, Chamcepelia talpacoti. B, Left wing of male Ch. passerina. C, Oblique bar in the "down'' stage in male Geopelia humeralis, age 16 days. D, Juvenal Geopelia humeralis 82 30. A, Wing of adult female Chamcepelia passt nun. B. Adult ' 'hamospelia buckli iji. C, Selected feathers from juvenal ( 'hamapi lia pallescens 82 31. A, Adult male Pvrislrra rim mi. 1-9, feathers from left wing of adult male /'. cinerea 84 32. Illustrating origin of oblique wing-bar in Columbula picui. A. Columbula picui. B, Lilt profile of adult Chamcepelia passerina. ('. Right wing of Ch. passerina si 33. Illustrating origin of oblique wing-bar in ( 'olumbula picui. A, Right wing of ' 'hama pt lia passt rina. B, Male Peristera mondetoura. C, Peristera sp.? D, Adult ma le /'< ristt ra geofroyi 84 34. A, Adult male inca-dove, Scardafella inea. B. Neck-feather of Sc. inca. ('. Wing of juvenal inca-dove, 3 weeks old. D, Scapular from juvenal Sc. iura SO 3o. A, Wing of adidt male inca-dove, Sc. nun. I',. Adult male ' 'halcopt lia afra 86 36. Adult scaled dove, Scardafella squamosa. 1-9, Selected feathers from an adult Sc. inca 8S 37. A, Adult male (imprint striata. Figs. 1 to 7, Separate feathers from normal G. Striata 90 38. A, Juvenal G. striata. B, Adult female a. striata 90 39. A, Adult Geopt lia Iranquilla. Figs. 1 to 9, Neck-feathers of adult G. Iranquilla 92 40. A. Adult male Geopelia humeralis. B, Feather from near front edge of neck of G. humeralis. C, Juvenal (,'. humeralis, age 28 days. D, Feather fr side of neck of G. humeralis. E, Feather from middle of breast of G. lawn rails 94 41. A, Wing of juvenal G. humeralis, age M days. B, Wing of juvenal G. humeralis, age I8days. < '. Four inner tertials of G. humi ml is. 1), Three inner tertials of Zenaidura carolinensis (PP1). E, Three inner tertials of Z. carolinensis (PP2) 94 42. A. Female cape-pigeon, Ot na capt nsis. B. Right wing of adult female tambourine-dove, Tympanistria tym- panislria. < ', Three inner tertials of juvenal female tambourine-dove. D, Two inner long coverts from same bird. E. Wing of adult male Chalcopi lia afra 98 viii ILLUSTRATIONS. IX FACING PAGE 43. Figs. 1, 2, 3, Three wing-feathers of adult female green-winged pigeon, Chalcophaps indica. Fig. 4, Posterior scapular of a male Phaps elegans. Fig. 5, From a female nicobar-pigeon, Calocnas nicobarica. Figs. 6, 7, S, Three wing-feathers of a juvenai (7 weeks) Leptoptila sp.? 100 I 1. Selected feathers from white-faced pigeon, Leucosarcia picala 102 45. Right wing of adult crowned pigeon, Goura coronala 104 46. A, Wing of Columba rufina. B, Wing of Columba unicincla 106 47. Adult Columba squamosa 106 48. A, Colored photograph of head and neck-mark of the wood-pigeon. C. palumbus. B, Colored photograph of head and neck-mark of C. gymnopthalma. C, Colored photograph of anterior parts of adult male. Geophaps smithi 108 19. Adult male and female crowned pigeon, Columba leucocephala 108 50. Juvenal Columba U uconota, age 4 weeks 108 51. A, Adult female common pigeon, No. K2. B, Adult female common pigeon, No. Kl 110 52. A, Adult female common pigeon, No. K3. B, Adult female homer 110 53. A, Juvenal male common pigeon, age 5 weeks. B, Juvenal common pigeon, age "> weeks 112 54. A, Adult male H Columba guinea X % Columba tabellaria hybrid (G7CC). B, Adult male C. guinea X homer hybrid (G7CCA). C, Adult male C. guinea X homer hybrid (G7< !CB) 112 55. Figs. 1-11, Advanced stages of the "guinea-mark" in scapulars and coverts of a juvenai common pigeon Fig. 12, Long covert of adult female satinette dove 1 1 li 56. Selected leathers from juvenai robin, Mcrula migratorius, age 3 weeks. First series 122 57. From juvenai robin. From upper side of same birds and in same order as pi. 50. Second series 122 58. Analogies of the robin pattern in common pigeons 124 59. Selected feathers from an adult male flicker, Colaptes auratus 126 60. A, The sharp-tailed sandpiper, Tringa acuminata. Figs. 1-11, Selected feathers from an adult quail, Colinus mrginian us 126 ill. Figs. 1-11, Selected feathers from an adult female ring-necked pheasant of China, Phasianus torqualus, . . . 126 62. A, The splendid parrakeet, Plalycercus sph ndidus. B, The curlew, Numt nius arquata 128 63. A, The black-backed hemipode, Hemipodius melanotus. B, The collared plain-wanderer, Pedionomus torqualus. C, Adult male European widgeon. Anas penelope 128 64. A, Juvenal crowned guinea-fowl, Numida coronala. B, Wing of adult .V. coronata 130 65. The ring-necked pheasant of China, Phasianus torqualus 130 66. Phasianus wallackii 132 67. Selected feathers from a male Simla horned pheasant, Tragopan melanocephalus 132 68. Polypleclron bicalcaratum 132 69. A, B, Third left primary of Columba guinea. C, D, Fourth left primary of common pigeon (mondain). E, Frilled wing-coverts in an adult female owl-roek hybrid 140 70. A, Adult male common pigeon (mondain). B, Adult male guinea-pigeon, Columba guima 142 71. A, Adult male common pigeon, red with white primaries and small guinea-marks. B, Adult female homer (E). C, Adult female homer (E), same as above 142 72. Transverse barring in an adult male owl-pigeon 151 73. Fundamental bars and extended wing-bars in adult male owl-pigeon . . 154 74. Fundamental bars in wing of juvenai Jacobin 156 75. Barred secondaries and long coverts of a Jacobin, age S weeks 156 70. Mate (19-normal) of guinea mutant (No. 20) at age of 5 months 162 77. C. guinea, No. 20, "mutant" 162 78. Mutant < '. guinea (20) and nest-mate, normal ''. guinea (19), age 26 days 164 79. A, Left wing of a juvenai pheasant, age 6 days. B-F, The inner greater coverts, left wing, of ' '. guinea anil other species of Columba. B, C. livia domeslica, adult. C, C. maculosa, adult. D, C. guinea, mutant (20) adult, age 14 months. E, C. guinea, normal (19) adult, age 11 months. 1'. C. guinea, normal juvenai, age 4 weeks I' 1 ' 80. Head, eyes, anil naked periorbital area of C. guinea, mutant and normal, and of other Columba. A. Adult C. guinea, normal (19). B, Adult C. guinea, same. < ', Adult C. guinea, mutant (20). D, Colored photograph of head and neck of C. eversmanni. E, C. domestica. F, C. leuconota 166 SI. A, Tail of Columba guinea, normal (19), age 70 days, li. Tail of C. guinea, mutant (20), age 78 da) s. < '., Tail of C. guinea (10), age 125 days Inli 82. Young (MN1) of Columba guinea mutant, age 1 weeks 166 S3. A, Adult male Zenaida vinaceo-rufa. B, Juvenal female Z. vinaceo-rufa (12), age about 8 weeks 170 si. A, Juvenal normal (20) Zenaida vinaceo-rufa, age about 8 weeks. B, Juvenal "mutant " (21 I Z. vinaceo-rufa, age 6 weeks ' ' " 85. Offspring of mutant Zenaida vinaceo-rufa (21). A. Z, naida mutant (21 1 age 6 weeks B, < '. First-generation hybrid, Z-ZN5. D, E. Two views of first-generation hybrid, Z-ZX1. F, Two third-generation mutant hybrids, 3Z-6ZN3 and t l72 86. A, B, C, Later offspring (1910) of the original Zenaida mutant (21 I. D, K. /■ naida mutant offspring of 1915 and 1916 'i 1 87. Adult female Zenaida amabilis ' 88. A, Juvenal Geopelia euueala (Gcl-A2), age 15 days. B. Juvenal (i. cuneata (Gel CI I, age 10 days. C, Juvenal G. cuneata (Gcl-Cl), age 32 days '92 X ILLUSTRATIONS. TEXT FIGURES. PAGE 1. Adult male barred rock-pigeon, Columba livia, No. 17 2. Adult male barred rock-pigeon, C. livia, No. 7 IS 3. Oriental or Japanese turtle-dove, Tartar orientalis 19 4. Adult female Florida ground-dove, Chamoepelia passcrina 24 5. Chequers of wing of juvenal male passenger-pigeon, Ectopisles migratorius 25 0. Chequers of wing of adult female passenger-pigeon 25 7. Chequers of adult male passenger-pigeon 27 S. Adult white-winged pigeon, Melopelia It ucoplera 29 9. Adult white-breasted crested pigeon, Lophophaps leucogaster 31 10. Juvenal male crested pigeon, Ocyphaps lophotes 33 11. Adult female crested pigeon, O. lophotvs 38 12. Diagram to illustrate pattern sequence in I he anterior wing-bar of male Ocyphaps 44 13. Adult male Nesopelia galapagoensis 56 14. Left wing of juvenal white-winged pigeon, Melopt lia It ucoplera, age 25 days 57 15. Showing the oblique streak in a common pigeon of 7 weeks and 2 days 79 16. Pin-feathers marking the region of the oblique streak in a common pigeon of 17 days 81 17. Adult Mange's dove, Geopelia maugei S7 18. Wing of juvenal Geopelia striata, age 18 days 89 19. Wing of juvenal G. striata, age 28 days 89 20. Wing of juvenal G. humeralis, age 30 days 93 21. Wing of juvenal G. humeralis, age 10 weeks 95 22. Wing of adult G. humeralis 95 23. Columba rupestris 110 24. A, The Philippine rail, Ballus pectoralis. B, The wood sandpiper, Totanus glareola 129 25. The breast-frill in j uvenile hybrids 144 26. A, Hybrid mourning-dove X ring-dove. B, Adult Geopelia striata. C, Same fraternity with three-fourths hybrid of text fig. 25. D, Juvenal Geopelia humeralis, just hatched 145 27. A, Juvenile, three-fourths hybrid, same fraternity with text-figure 25. B, Juvenile homer (C2) in first plumage, age 06 days. C, Adult male chequered homer 1 16 28. Three stages of frill and feathering in a juvenal African owl-pigeon. A, African owl, age 17 days. B, The same, age 26 days. C, The same, age 36 days 147 29. A, Adult male African owl-pigeon. B, Adult hybrid Zenaidura X St. risoria. C, The same, in profile .... 148 30. Juvenal Geopelia tranquiUa (Gtl-Bl), age 10 days 186 31. The same, age 17 days 187 32. Juvenal G. tranquiUa (Gtl-Bl), age 2S days 18S 33. The same as text-figure 32, age 38 days 189 34. Geopelia cuneata, age 68 days (second plumage) 191 35. Juvenal G. cuneata. age 44 days 193 36. Geopelia cuneata (Gel -Al ) . age 68 days 193 ERRATA. Page 73, line 36, for "pis. 16 and 17, Vol. II," read "pi. 8, Vol. II." Page 76, line 5. for " Fig. B. 5, 0. 7, 8," read " Fig. B. 1. 2, 3. 4." Page 76, line 11, for "Fig. D. 1. 5. 6." read " Fig. D. 1. 2, 3." Page 80, line :!, for "C. pigui" read "C. picui"; also on page 84. Page 84, under explanation of Plate 32, Fig. B. For "X 0. 5" read "natural size." Page 94, explanation of Plate 40, under Figs. A and C, instead of "X 0.7" read L.4. Page 107, line 23. for "pi. r,7. Vol. II." read "pi 35, Vol. II." Page L28, near foot of page, for "X 0.75," read "X 0.35 (about)." Plates 70 ami 80, periorbital is misspelled. ORTHOGENETIC EVOLUTION IN PIGEONS POSTHUMOUS WORKS OF CHARLES OTIS WHITMAN EDITED BY OSCAR RIDDLE VOLUME I CHAPTER I. INTRODUCTORY. 1 Progress in science is better indicated by the viewpoints we attain than by- massive accumulation of facts. Darwin's perspective made him a prodigy in the assimilation of facts and an easy victor in the greatest conflict that science has thus far had to meet. His triumph has won for us a common height from which we see the whole world of living beings as well as all inorganic nature; phenomena of every order we now regard as expressions of natural causes. The supernatural has no longer a standing in science; it has vanished like a dream, and the halls consecrated to its thraldom of the intellect are becoming radiant with a more cheer- ful faith. In no other field of science is the viewpoint of such ever-present importance as in biology. It means orientation of the field, clear vision in directing research, and insight in the interpretation of phenomena. It may be well, therefore, to prepare the ground a little in advance of the facts to be presented, and this can be done to best advantage by taking our bearings from two of our foremost guides — Darwin and de Vries. The great problem in biology, as it presented itself to these leaders, was to account for progressive evolution by natural means. Supernaturalism had declared the living world a stupendous miracle, a creation by fiat. Science had already made good its claim to the physical world, but the animate sphere was still supposed to rise supreme above natural law. Darwin was taught the creation dogma, and all through the memorable voyage of the Beagle he was predisposed to regard species as immutable. He stated in a letter to Dr. Otto Zacharias, in 1877 : 2 When I was on board the Beagle I believed in the permanence of species, but, as far as I can remember, vague doubts occasionally flitted across my mind. On my return home in the autumn of 1836, I immediately began to prepare my journal for publication, and then saw how many facts indicated the common descent of species, so that in July, 1837, I opened a note-book to record any facts which might bear on the question. But I did not become convinced that species were mutable until, I think, two or three years had elapsed. The five-year voyage on the Beagle, during which facts were collected with no definite theory but with a haunting doubt as to the immutability of species, was the foundation or germ-stage in the development of Darwin's great generalization. It is still a notable fact that the germ of Darwin's views at this inceptional stage was already a logical whole so far as fundamentals were concerned. The problem embraced the entire organic kingdom, from lowest to highest. The great difficulty, then, was to find out how nature could be self-regulating in bringing forth such nice adaptations as are everywhere patent in the organic 1 The manuscript here selected as introductory to this volume was written in 1909, and formed part of a lecture at Clark University, entitled "A study in evolution, based on color-characters in pigeons, and bearing on moot ques- tions." A few phrases of that address are omitted. The data presented in connection with this lecture were not reduced to writing, but accompanying lists of topics, illustrations, etc., show that many of the subjects which are fully treated in the several chapters of this volume were diseased. — Editor. ' Cited by Huxley in his Darwiniana Essays, p. 276. 3 4 ORTHOGENETIC EVOLUTION IN PIGEONS. realm. How could teleological results flow from non-teleological causes? That was the staggering problem that confronted Darwin and kept him from feeling entirely convinced "two or three years" longer of the mutability of species. Already he had "perceived that selection was the keystone of man's success in making useful races of animals and plants; but how selection could be applied to organisms living in a state of nature remained for some time a mystery." 3 For clearing up this mystery the credit still belongs to Darwin, for he alone knew there was such a mystery and made the application that dissolved it. As Huxley remarks: 4 Variation occurs under natural, no less than under artificial, conditions. Unrestricted multiplication implies the competition of varieties and the selection of those which are relatively best adapted to these conditions. This is a concise statement of the theory of natural selection, so simple and self-evident that it at once carries conviction. The theory tells us that nature is self-regulating in the living as in the non-living world. It tells us that the more useful variations have the best chance of survival in the struggle for existence, and thus shows that, variation being given, progressive evolution is inevitable. The theory does not undertake to explain the first appearance of life or to state the causes of variation, but it does show how nature sifts out the fit from the unfit and builds upon such variations as have a winning value in the battle of life. For this all-embracing viewpoint, which completes and crowns the advances by Newton and Lyell, the world is indebted to Charles Darwin. The theory of de Vries, the distinguished author of the mutation theory, coincides in the main with that of Darwin; but he introduces a distinction in regard to variations which he regards as fundamental, dividing them into two classes: (1) ordinary or fluctuating variations, and (2) mutations. All species exhibit "fluc- tuations," but these obey Quetelet's law of probability and never transgress the limits of the type. From them improved races may result, but never new species. Mutations, on the other hand, are those variations which remain constant, and these are the sole source of new species and new varieties. Fluctuations are regarded as merely quantitative, while mutations stand for qualitative changes. However small or large, mutations are always sudden in appearance; that is, they do not arise by slow and transitional degrees, but come ready-made; and although they are subject to fluctuations, they are essentially immutable units — the so-called "unit-characters." A species thus consists in some definite number of these fixed units, from the germ onward to the full adult stage, and this number remains con- stant. The loss or addition of a single unit would make at once a new species. The fundamentals in the mutation theory are: 1. Every species consists of a fixed number of unit-characters. 2. The species and the component units are alike sudden in origin and unchanging in type. 3. Old characters may be suddenly transmuted into new ones, but between the two there is always a gulf of absolute discontinuity, with no possible bridge of modification. 4. Continuous intergradations may connect a species with an improved race, but never one species with another species. • Life and Letters, I, p. 68. * Darwiniana Essays, p. 279. INTRODUCTORY. 5 5. Species-formation is chancewise in direction, never resulting from a tendency to vary in any one determinate direction. 6. Natural selection can not give origin to new species; it can only weed out from those already in existence such as are incapable of sustaining themselves. The central conception of this whole theory of mutation is that of fixed unit- characters. Specific characters are called units, because they are supposed to be individualities and to remain such in all stages of development and evolution. If such a character undergoes change, it ceases to be the same character and by sudden transformation becomes an entirely new unit, or unit-character. Trans- form one of these units, add one, or remove one, in each case the old species becomes a new species per sal turn. These units are imagined to have the clean-cut individ- uality of a chemical molecule, and any real change is likened to a chemical substitution. Such views are captivating, for they offer definiteness in place of the vague and mysterious. The problem of the origin of species hangs on the nature and meaning of specific characters. Reveal the secrets in the development and evolu- tion of a single specific character, and you will furnish the key for understanding all characters and all species. Right or wrong, deVries has done well to direct attention to the all-important point, the specific character. Here, I believe, lies the whole problem of evolution, and de Vries is right in trying to draw a line between specific and non-specific variations. It should not be overlooked, however, that Weismann has already set us far along in this task; for not long ago he convinced most of us that only germinal variations are hereditary and that all so-called "acquired characters" of somatic origin are non-transmissible and therefore without specific significance. Do we not have, then, in germinal variation, a better criterion of what is specific than we get in sudden appearance? Indeed, is it not here that the seeming sudden- ness of first appearance finds its explanation, and likewise the fact that so-called mutations involve the whole organism? If we are to accept the physiological con- ception of development, as is inevitable in my opinion, it is easy to see that a change, however slight, in the primordial constitution of the germ would tend to correlate itself with every part of the whole germ-system, so that the end-stage of development would present a new facies and appear as a total modification, answer- ing to what de Vries would call a mutation. That something of this order does sometimes occur I have indubitable evidence, and in such form as to dispel the idea of discontinuity and sudden gaps in transformation. The idea of unit-characters, however, as distinct elements that can be removed or introduced bodily into the germ does not appeal to me as removing difficulties, but rather as hiding them; in short, as a return to the old pangenesis view of preformed characters. In this theory, as is well known, we had two miracles in- volved. The first consisted in a centripetal migration of preformed gemmules and the second in the centrifugal distribution of the same elements. De Vries dismisses the first of these, but accepts the second, and on it rears the superstructure of his theory of mutable-immutable unit-characters. With all due respect to the dis- tinguished author of this theory, and with abounding admiration for his great work and model methods, which have aroused universal interest and stimulated enor- mously experimental bionomics, I am strongly persuaded that his hypothesis of unit-characters fails as a guide to the interpretation of the species and its characters. C ORTHOGENETIC EVOLUTION IN PIGEONS. It is true that a great amount of work on Mendelian heredity seems strongly to support the unit-character hypothesis, and that cytology offers some further support. Nevertheless, I have to confess to wholesale skepticism. The germ, as I believe and have long maintained, stands for an organized whole. It is a unit organ- ism, not an organism of units; all the features that arise in course of development are within the sphere of the individual unity and integral parts of it, and whatever specificity they possess is completely determined and not of independent origin. The strongest suggestion of unit-characters is found in the Mendelian phenome- non known as segregation. I do not underestimate the importance of this striking behavior of so-called alternative unit-characters. I am familiar with them, and deeply interested; but I am unable to see in them the sum total of all we know about heredity. What J have said in regard to unit-characters applies to the Mendelian doctrine. Mendelism, like mutation, neglects the natural history of the characters it experiments with and is not primarily concerned to know how characters have originated and multiplied. It seems to me a great error for the mutationist and the Mendelian to construe characters as disconnected entities rather than as modifications slowly and gradually evolved, in genetic continuity from stage to stage, as we see so well illustrated in normal development. De Yries attaches considerable weight to the fact that he has actually seen mutations and knows their pedigree for one or a few genera- tions. Since the publication of his remarkable work, other investigators have reported numerous mutations, and I must join this group to the extent of admitting that I have witnessed phenomena that a mutationist would probably claim as confirmations of his theory. Some of these phenomena require a very detailed and specific treatment; but let us here define the nature of the field we are to enter and make clear the way of approach. The ability to interpret evolutional phenomena, or even fully to understand a given interpretation, implies not only concentrated attention but also a prepared state of mind. I wish, therefore, first of all, briefly to discuss a query that may obtrude itself at the outset. The simple lifeless color-mark of a feather may, at first thought, appear to be artificial, too extremely variable, to deserve serious attention. But is not variation the foundation of all evolution? and is it not a great advantage to have our subject-matter where it can be easily seen? If the very exuberance of these colors and color-patterns be somewhat appalling on first approach, it costs but little effort to see that it all counts in favor of the investigator. The greater the variability, the closer, in general, will be the connections between stages, and the easier it will be to catch the trend of derivation, and to discover the common points of departure for whole groups of related color-patterns, and, pos- sibly, to reduce these points to a single point of departure for the whole bird king- dom. Such a vista once opened would orient the whole field, disclose the direction or directions of evolution, provide the investigator with a key to the natural order of sequence in color-patterns, enable him to detect and to demonstrate orthogenetic evolution, if such there be, and to discriminate nicely between this and the results to be ascribed to natural selection and other intervening factors. From such a vantage-ground, juvenile stages of color-patterns would become luminous as re- capitulations, in the sense of the biogenetic law, and not stand as isolated prodigies of mutation. INTRODUCTORY. 7 Nature has a long memory in that as she has built slowly through the ages, so she now rebuilds in each individual development, repeating, over and over, in each generation, the essential steps taken in the preceding generation. This drama of recapitulation, in every stage a tableau vivanl, condensed and modified as of neces- sity it must be, since every stage is to some extent a remolding of the earlier, is nature's silent rehearsal of past history, or what is commonly called heredity. This is unquestionably the greatest wonder that biology has yet disclosed. Its fundamental significance has scarcely yet found general appreciation even among students of evolution. A few — Hering, Haeckel, Semon, Francis Darwin, and especially Rignano — have grasped and treated its deeper meaning as a universal mnemonic law, a law that underlies all intelligence as well as all development. The disputations hitherto held over Haeckel's biogenetic law do not as a rule touch the essential phenomenon; indeed, they obscure the marvelous accuracy of hereditary recapitulation. This uniformity of nature's laws and the perpetual repetition of her building processes are the two basal facts on which rests the hope of all cumulative science and of our ultimate triumph in reading the secrets of evolutionary history. These remarkable phenomena, which in normal development furnish, as a rule, only frag- mentary and disconnected parts of past evolutional history, may be so expanded, by suitable experiments, as to demonstrate complete continuity of stages in the passage to the adult pattern. This crucial experiment consists in plucking a few juvenal feathers at such points and time-intervals that the new feathers developing in the places of those removed will appear at successively later ages, each unfolding a pattern in a stage of evolution corresponding to its age. The result is that we get an ascending series of stages rising gradually from the juvenile to the adult pattern, so that we have pictured before our eyes the progressive transformation by which the earlier is converted into the later and final pattern. One such experiment as this reveals a law that must hold, in principle at least, for the pigeon phylum and all its branches; and not only for this group of birds, but also for the entire bird world. Of course it is not to be expected that such an experiment would give equally conclusive results in all cases, for different species have run different courses and the length and fulness of the recapitulative phases vary greatly. But all develop- ment, it must be remembered, is essentially a repeating or recapitulating process. This is the central fact of heredity and the doctrine of descent. The first or germ stage is, of course, the oldest in the phyletic series, and the adult stage, which we commonly call the older stage, considering the age of the individual, is really the youngest of the phylum. From this point of view it is readily seen that we are to expect the fullest recapitulation in the final stages and the most condensed or abbre- viated recapitulation in the initial stages. The development of the individual is everywhere a confirmation of this anticipation, if we allow for the fact that certain stages have greater physiological value than others, and consequently have been favored and kept at the level of need. 6 The cell-stage of the germ is an instance. 6 This is the work of natural selection, which is omnipresent and efficient in every part of the organic universe. To call for proof of the action of selection is to call for proof of such obvious facts as the struggle for existence and the survival of the fittest. 2 8 ORTHOGENETIC EVOLUTION IN PIGEONS. When, then, an evolution of millions of years has to be repeated in a few days or weeks, there must be enormous abbreviation in earlier stages, relieved here and there by a few essential landmarks, and only in the later stages relatively fuller exhibition of details. If we are fortunate enough to find a case here and there where lost history can be recalled and actually pictured forth in the living bird, we have, as I have said, a demonstration of a physiological law in development that must have universal validity. The process is not less one of recapitulation or repetition for being condensed or abbreviated. These surface-characters, then, that at first sight seem so bewildering in their profusion, so baffling in their sudden seasonal and nuptial transmutations, so puzzling in their age-sequences, turn out to have quite peculiar advantages for the student of evolution. Their perfection as specific marks, delicacy as tests of variability, accessibility in breeding and hybridizing experiments, the convenient serial disposition of their intergradations in rows of feathers, the phyletic sequences they reveal in successive plumages, the ease with which we can force them to bridge gaps in normal development and reproduce lost stages, all these are advantages that speak for themselves. In some birds there are several successive patterns graduating up to the adult pattern through as many molts, partial or total. In such cases it is often possible to find practically all the transitional phases that connect the earlier with the later pattern. Here nature turns the molting process into an experiment essentially the same as that I have tried on the diamond-dove and other geopelias. Whenever a bird develops its feathers slowly, so that they appear at different ages in different regions, as in the jungle-fowl, quails, pheasants, etc., it is easy to see that the pas- sage from stage to stage is not by mutations, but by continuous development. The testimony from the avian world in respect to the evolution of color-patterns is, as I have said, often fragmentary, and herein is to be found the only apology for an appeal to the theory of mutation. Wherever the evidence is fairly complete and wherever we find it possible to bring the apparent breaks in continuity of development to a crucial test, there we find positive confutation of the mutation hypothesis. The evidence for mutation is mainly of a negative character. New species arise, and we discover them when it is too late to learn the history of their genesis. We search through the biological world for discontinuities, and we find them so numerous that ponderous volumes can not approach a full record of them. In our breeding- experiments, with both animals and plants, we not infrequently get departures from the parent type, and these unexpected novelties, seemingly in contradiction to the law of hereditary continuity, are then declared to have come suddenly into existence, to have slipped the bonds of parentage, to have hoisted themselves into existence inconsequentially, persaltum. Shall we assume a break in the chain at every point where our knowledge fails? or is it more rational to abide by the law of genetic continuity so long as we fail to discover evidence of an exception? CHAPTER II. THE PROBLEM OF THE ORIGIN OF SPECIES. 1 Among the rival theories of natural selection two are especially noteworthy. One of these is now generally known as orthogenesis.- Theodore Eimer was one of the early champions of this theory, basing his arguments primarily upon his re- searches on the variation of the wall-lizard (1874-81). Eimer boldly announced his later works on The Origin of Species (1888) and the Orthogenesis of the Butter- flies (1897) as furnishing "complete proof of definitely directed variation, as the result of the inheritance of acquired characters, and as showing the utter impotence of natural selection." Eimer's intemperate ferocity toward the views of Darwin and Weismann, coupled with an equally intemperate advocacy of the notion that organic evolution depends upon the inheritance of acquired characters, was enough to prejudice the whole case of orthogenesis. Moreover, the controversial setting given to the idea of definitely directed variation, without the aid of utility and natural selection, made it difficult to escape the conclusion that orthogenesis was only a new form of the old teleology, from the paralyzing domination of which Darwin and Lyell and their followers had rescued science. Thus handicapped, the theory of orthogenesis has found little favor outside the circle of Eimer's pupils. The second of the two theories alluded to is the mutation theory of Hugo de Vries. The distinguished author of this theory maintains, on the basis of long- continued experimental research, that species originate, not by slow, gradual variation, as held by Darwin and Wallace, but by sudden saltations or sport-like mutations. According to this theory, two fundamentally distinct phenomena have hitherto been confounded under the term "variation." In other words, variation, as used by Darwin and others, covers two classes of phenomena, totally distinct in nature, action, and effect. Variation proper is defined as the ordinary, fluc- tuating, or individual variation, and this is held to be absolutely impotent to form new species. It is claimed that no amount of either natural or artificial selection can by any possibility lead this variation up to the birth of a new species. The utmost that could be attained would be an improved race that would inevitably revert to the original state as soon as left to itself. Mutation, on the other hand, never advances by slow and minute modifications, which are continuous and cumulative, but by single, sudden jumps. In the words of de Vries (vol. 1, page 150) : Species have not arisen through gradual selection, continued for hundreds or thousands of years, but by jumps [stufenweise] through sudden, though small, transformations. In contrast with variations which are changes advancing in a linear direction, the transfor- mations to be called mutations diverge in new directions. They take place, then, so far as experience goes, without definite direction. 'An address before the Section of Phylogeny and reprinted from "Congress of Arts and Science, Universal Exposition, St. Louis, 1904," Vol. V. A few introductory paragraphs have been omitted, and a very few minor addi- tional changes have been made by the editor in adapting the address to this position in this volume. 2 A name introduced by Wilhelm Haacke (Gestaltung und Vererbung, p. 31). 9 10 ORTHOGENETIC EVOLUTION IN PIGEONS. The new species arises from the old, but without any visible preparatory steps and without intermediate connecting stages. Like the old, it is subject to varia- tion, but as a type it is essentially immutable. Do Vries does not deny that varia- tion produces what may appear to be transitional forms, but he maintains that these forms in reality have no such meaning. They are to be regarded as phenomena of " transgressive variability," which may obscure but not obliterate the specific limits. De Vries states (vol. 1, page 362) : The transitions do not appear before the new species, at most only simultaneously with this, and generally only after this is already in existence. The transitions are there- fore not intermediates or preparations for the appearance of the new forms. The origin takes place, not through them, but wholly independently of them. Granting that the position with respect to the mutants obtained from the evening-primrose (Oenothera lamarckiana) is unassailable, does it follow that all new species have arisen by mutation and that continuous variation has never had, and never can have, anything to do with the origin of species? Plausible as is the argument and impressive as is the array of evidence presented, I can but feel that there are reasons which compel us to suspend judgment for a while on this pivotal point of the mutation theory. America is the original home of the evening-prim- roses, and it is here that the natural history of the group remains to be worked out in the light of the experimental results obtained in Holland. What does it mean that a few' mutants keep on reappearing year after year, and that even the mutants themselves mutate, not in new lines, but in the same old ones? Persuaded as deeply as I am that we can never draw from a species anything for which no ancestral foundations preexist, I anticipate that our wild evening- primroses have revelations to make. 3 Whatever revelations may await future investigation in this field, the work done in the primrose garden of Amsterdam will stand as a classical contribution to the new biology and as one of the very best models in method of research that we have yet seen. Natural selection, orthogenesis, and mutation appear to present fundamental contradictions, but I believe that each stands for truth, and that reconciliation is not distant. The so-called mutations of Oenothera are indubitable facts; but two leading questions remain to be answered: First, are these mutations, now appearing, as is claimed, independently of variation, nevertheless the product of variations that took place at an earlier period in the history of these plants? Secondly, if species can spring into existence at a single leap, without the assistance of cumulative variations, may they not also originate with such assistance? That variation does issue in new species, and that natural selection is a factor, though not the only factor, in determining results, is, in my opinion, as certain as that grass grows, although we can not see it grow. Furthermore, I believe I have found indubitable evidence of species-forming variation advancing in a definite direction (orthogenesis), and likewise of varia- tions in various directions (amphigenesis). If I am not mistaken in this, the reconciliation for natural selection and orthogenesis is at hand. » This was written (1904) before the publications of Davis and of Jeffrey on the (Enoihera.— Ed. THE PROBLEM OF THE ORIGIN OF SPECIES. 11 I am aware that orthogenesis is held by many to be utterly incompatible with both natural selection and mutation. De Vries says (vol. 1, page 40): The Darwinian principle demands that species-forming variability and mutability be indeterminate in direction. Deviation in all senses must arise, without favor to any par- ticular direction, and especially without partiality for the direction proceeding from the theory to be explained. Every hypothesis which departs from this principle must be rejected as teleological, and therefore unscientific. Again (page 180) the same point is amplified: Again and again, and by authors of different aims, it has been insisted upon that species-forming variability must be orderless. The assumption of a definite variation- tendency which would condition, or even favor, the appearance of adaptive modifications lies outside the pale of the natural science of to-day. In fact, the great advantage of Dar- win's doctrine of selection lies in this, that it strives to explain the whole evolution of the animal and plant kingdoms without the aid of supernatural presuppositions. According to this doctrine, species-forming variability goes on without regard to the qualification of the new species for maintaining themselves in life. It simply supplies the struggle for existence with the material for natural selection. Whether this selection takes place between individuals, as Darwin and Wallace supposed, or decides between whole species, as the mutation-theory demands, ultimately it is, in either case, simply the ability for existence under given external conditions that decides upon the permanence of the new form. I take exception here only to the implication that a definite variation-tendency must be considered to be teleological because it is not "orderless." I venture to assert that variation is sometimes orderly and at other times rather disorderly, and that the one is just as free from teleology as the other. In our aversion to the old teleology, so effectually banished from science by Darwin, we should not forget that the world is full of order, the organic no less than the inorganic. Indeed, what is the whole development of an organism if not strictly and marvelously orderly? Is not every stage, from the primordial germ onward, and the whole sequence of stages, rigidly orthogenetic? If variations are deviations in the directions of the developmental processes, what wonder is there if in some directions there is less resistance to variation than in others? What wonder if the organism is so balanced as to permit of both unifarious and multifarious variations? If a developmental process may run on throughout life (e.g., the lifelong multiplication of the surface- pores of the lateral-line system in Amid), what wonder if we find a whole species gravitating slowly in one or a few directions? And if we find large groups of species all affected by a like variation, moving in the same general direction, are we com- pelled to regard such "a definite variation-tendency" as teleological, and hence out of the pale of science? If a designer sets limits to variation in order to reach a definite end, the direction of events is teleological; but if organization and the laws of development exclude some lines of variation and favor others, there is certainly nothing supernatural in this, and nothing which is incompatible with natural selec- tion. Natural selection may enter at any stage of orthogenetic variation, preserve and modify in various directions the results over which it may have had no pre- vious control. It should be noted in this connection that the difficulty of incipient stages is not what it is so generally presumed to be. The advocates of natural selection 12 ORTHOGENETIC EVOLUTION IN PIGEONS. habitually assume that the evolution of an organ or character begins with an ''infinitesimal rudiment," which has no way of emerging from its functionless state except through minute chance variations in various directions. In this assumption the problem is misconceived. The characters we meet with to-day have rarely, if ever, arisen by direct evolution from useless rudiments. When we know enough about a character to undertake to trace its genesis, the "rudiment" imagined to lie so near recedes, and we are led on, not to a "beginning," but to an antecedent, and if we are fortunate enough to be able to advance farther, we come to another antecedent, and so on. The series of antecedents stretches ever as far as we can see. As we repeat this experience with different characters, looking always for the primordial rudiment, our childish faith in such "beginnings" gives way to the conviction that the chase is led by a phantom. No one of our sense-organs, for example, can be traced to a rudiment, except in the embryological sense. The eye of the vertebrate may appear as a rudiment in the embryo, but no one can doubt that it has had a phylogenetic history, the first term of which — if first there be — must have been very different from its present embryonic rudiment. To assume that the eye began in some different variation that fluctuated or mutated, chance-wise, into a state of incipient utility, and was then developed in a direct line to its present stage of complex adaptations, either gradually or per saltum, would be hardly more satisfactory than appealing to a miraculous succession of miracles. It is impossible to believe that such a system of harmonious coadaptations could ever arise by mutation; 4 and selection, although playing a very important part in such achievements, is probably never equal to the whole task. Without the assistance of some factor having more continuous direct- ive efficiency, selection would fail to bring out of the chaos of chance variation, or kaleidoscopic mutation, such progressive evolution as the organic world reveals. In order to show that such a factor is essential, and that it is actually present, supplying the indispensable initial stages and holding the master hand in the general direction of evolution, demonstrative evidence is, of course, required. Such evidence lies in the history of specific characters. But how shall we approach such a task, if no near-by rudiment is to be found as a starting-point? Rudiments and premutations are alike illusory in this regard, for their beginning is always and necessarily assumed to lie in the realm of the invisible and unknowable. If we are to keep always on ground that is open to investigation, we must find our starting- points in known stages. As the laws of nature are constant, it is not essential to trace entire histories. If some chapters are sufficiently open to observation and experiment to permit close study, we may hope, in some of the more favorable cases, to read the phenomena in their natural order and to learn from what goes on in one part of the history the factors that govern in all parts. The study of the problem of the origin of species resolves itself, therefore, ever more clearly into exhaustive studies of single favorable characters, in the more accessible portions of their history. For decisive evidence we must have characters of a comparatively simple nature, the evolutional records of which, in every case, are to be read in a considerable number of species of known common origin. 1 Darwin frequently emphasized the same objection. In a letter to Asa Gray, referring to the orchids, he remarks: "It is impossible to imagine so many coadaptations being formed, all by a chance blow." Weismann has shown in a masterly manner how inadequate is the mutation theory to account for such phenomena. B. \dult male two-barred rock-pigeon, Cdumba livia. X 0.4. Nov L909. From tteCaves of SoZty, Scotland, 1908 This bird has a * Ad^SSe'SSSd wild rock-pigeon, C. affinis. X 0.4 \. rding to , tSS only a variety of C. tfrfa. Nov. 1909. From Cav, r urty, Scotland, (1908). THE PROBLEM OF THE ORIGIN OF SPECIES. 13 It is a great mistake to resort exclusively to domestic races, for here the ancestry contains so many unknown elements that it is often impossible to refer phenomena to their proper sources. Even the so-called "pure breeds" are decidedly impure as compared with pure wild species. The ideal situation, as regards material, is to have pure wild -species in abundance as the chief reliance, and allied domestic races for subsidiary purposes. The pigeon amply fulfils all these prerequisites. A simple and convenient character, presenting divergent courses of evolution in some species and parallel courses in others, is to be found in the wing-bars and their homologues. It is to some chapters in the history of this character that we may now turn for evidence that natural selection waits for opportunities, to be supplied, not by multifarious variation or orderless mutation, but by continuous evolutional processes advancing in definite directions. The rock-pigeons (Columba livid) present two very distinct color-patterns, one of which consists of black chequers (pi. 1) uniformly distributed to the feathers of the wing and the back, the other of two black wing-bars on a slate-gray ground (pi. 1). These two patterns ma}' be seen in almost any flock of domestic pigeons. The inquiry as to the origin of these patterns involves the main problem of the origin of species, for the general principles that account for one character must hold for others, and so for the species as a whole. Darwin raised the same question, but did not pursue it beyond the point of trying to determine which pattern was to be considered original and how the derivation of the other was to be understood. Darwin's explanation was so simple and captivating that naturalists generally accepted it as final. It is but fair to state that Darwin's conclusions did not rest on a comparative study of the color-patterns displayed in the many wild species of pigeons. Accepting the view generally held by naturalists, that the rock-pigeons must be regarded as the ancestors of domestic races, the question was limited to the point just stated. It was known that the two types interbreed freely under domestication, and it had been reported that chequered pigeons sometimes appeared as the offspring of two-barred pigeons. Moreover, Darwin discovered that the chequers were homolo- gous with the spots composing the bars. As the main purpose was to show that variation was present to any extent required for the origin of new species, rather than to trace its course in any specific case, and as variation was supposed to be multifarious and progress to be guided by natural selection of the "fittest."' it is not strange that Darwin failed to get the direction of variation or to realize that in direction is given the key to one of the fundamental laws of evolution. As the two color-patterns are alike in having a common element, and differ chiefly in the number of elements, it was natural enough to take the smaller number as the point of departure and to view the larger number as "an extension of these marks to other parts of the plumage" (Animals and Plants, vol. 1. p. 225). With the ancestral type thus determined, and a simple mode of variation pointed out, Darwin could dismiss the problem with these words: "No importance can be attached to this natural variation in the plumage." Whence and how the two bars arose was not explained. The mode of depart- ure assumed to account for the chequered variety would, however, suggest that the bars themselves originated in the same manner; that is, from one or two spots 14 OHTHOGENETIC EVOLUTION IN PIGEONS. Explanation of Plate 2. Adult Columba livia, uniform gray, third bar. Photographed and spots measured Nov. 1898. There are 14 secondary and tertial coverts (primaries, 10). Left third bar (wholly concealed); measurements in millimeters. Feather 14. A blackish spot on upper edge and just a last trace of one on lower edge. Spots dusky gray; greatest width of spot, mm. long by 1.5 mm. 13. Upper spot, 6 mm. by 1.5, about same as in 14; lower spot so reduced that a magnifying glass is necessary in order to see the slight touch of dusky at lower edge. 12. Upper spot reduced to last touch, recognizable only by aid of magnifying glass; lower spot absent. 11. None on either side. 10. No upper; lower spot, 3 by 1. 9. Spot (lower), 3 by 1.5. 8. Spot (lower), 3 by 1.75. 7. Spot (lower), 5.5 by 3.25 (widest and most conspicuous of all). 6. Spot (lower), just a dusky trace. 5. Spot (lower), 5.5 by 1.5. 4. Spot (lower), 7.5 by 1.5. 3. Spot (lower), just a trace. 2. Spot (lower), no trace. 1. No trace. A third bar, wholly concealed, in this pigeon is suggestive. It is much the same as I have several times found in the first plumage of some young domestic pigeons. The rock-pigeon of (uniform gray wing) then exhibits in adult plumage the juvenile condition of domestic pigeons. Note, too, that two upper feathers exhibit double spots, at least in traces. The upper spot diminishes down- ward — is lost on No. 11. Of the lower spot there is only the minutest trace on two upper feathers; it is absent on the next two; and then, as the only spot, it increases downward, with a maximum on No. 7. The larger spots are in lower half of series, and this agrees in a general way with the young domestic. Notice for comparison chequered domestic (pi. 5), where the feathers of the third bar all have two spots instead of one. Feather No. M easurernents. Remarks. 17. Upper, lower 12 long by 7.5 End notch of gray, 2.5 deep (long) by 3.5 wide. 16. Upper, lower 14 5 long by 9.5 Gray tip, 4 mm. long, whole width of feather, nearly straight across. 15. Upper, lower 16 long by 10 Upper runs a little obliquely outward; gray tip, 6 mm. long, whole width of feather. lir long by 10 /Upper projects 1.5 mm. beyond lower. [Gray tip, 8 long, whole width of feather. 14. < ( [b Gray tip (upper), 10 mm.; upper spot 13. • paler, basal half largely gray. Lower 17 long by 12 Gray tip (lower), Li nun.; lower ob- liquely cut on proximal part. 12. • 11. 1 rpper absent.. }16 long by 12 (Gray tip, 16.5 mm.; obliquely cut on \ proximal part. Gray tip, 17 mm.; less obliquely cut on 17 long by 12 proximal part. 10. 9. 17.5 long by 1:2 18.5 long by 1-' Gray tip, 17 mm.; about same as No. 11 Gray tip, 17 nun. ; about smir as No. 11. 8. Lower 18 long by 12 Gray tip, 17.5 mm. ; spot begins to pale. 7. Lower 17 long by 11 Gray tip, 19 mm.; spot paler, mere diffuse on proximal part. 6. Lower 14.5 long by 10.5 Gray lip, 19 mm.; spot diffuse <>n both ends; longest on lower edge; inclines to point as it begins t" '1" in No; 7, and does more and more below. 5. Lower 14 long by 8.5 Gray tip, 20 mm.; spot begins to lit rent from shaft in No. 7, and retreats more and more downward. Spot is now considerably freckled with gray, and obliquely in both directions; more pointed. 4. Lower 11 long by 6.5 Gray tip, 20 mm.; freckled triangle. 3. Lower 7 long by 2.5 A freckled trace. 2. Lower 3 long by 1.5 Slill more diffusely freckled. 1. Lower No trace left. \ 4 Adult CoLumba lima, uniform gray, third bar. Photographed and asured Nov. 1898 THE PROBLEM OF THE ORIGIN OF SPECIES. 15 arising de novo, as chance variations, and the gradual extension of like spots in two rows of feathers. The one or more original spots, according to the general theory, would first appear as minute rudiments, and then be gradually enlarged and intensi- fied by the aid of natural selection, guided by their utility as recognition marks. Such a mode of origin would presuppose a plain, uniform gray ancestor, without any spots or bars in the wings, and would raise many puzzling questions that would be beyond the reach of investigation. For example: Why two bars? Why at the posterior end of the wing? Why do the spots taper backwards to a more or less sharp point in the chequered variety, while presenting a nearly square form in typical bars? Why should they have first extended upward, or downward, and in two rather than any other number of rows of feathers? If two rows of feathers were favored long enough to establish the bars for ornamental or other purposes, what freak of natural selection could have then interposed to turn a long-favored, definitely directed extension into a diffuse general extension, and thus to neutralize completely the very effects it was invoked to explain? Natural selection could not be supposed to originate or to guide the first indiffer- Explanation of Plate 2 — Continued. Notice. — (1) Double spots on 5 feathers. The upper spot disappears from the proximal end most rapidly on mesial or shaft side. Asymmetry on Nos. 14 and 13. (2) The lower spot follows the upper in losing ground on its proximal end, and mainly on mesial side, and thus keeps up a certain sympathetic change of form. But as soon as the upper disappears the lower fills up again on its near end and becomes nearly square. (3) The lower spot is most perfect on 5 feathers (Nos. 12 to 8), pales gradually from No. 9 downward, and loses in size rapidly from No. 7 downward. (4) The upper feathers have highest development, as shown in double spots, but the lower spot is best developed in the middle feathers — from No. 12 to No. 8. From No. 7 to No. 2 the spot wanes and disappears. Left first bar. — Spots of 14 tertials and secondaries. Feather No. Measurements. Remarks. 14 < ' Upper Il2.5 by 6.25 /Equal and symmetrical. I Lower J {Gray tip; broad notch, 2 mm. deep. 13. < ' Upper i Lower. 19 22 by by 6 9 ,Gray tip; broad notch; equal both sides. 12. < Upper 23 by h /Gray tip; 2 mm. above, 3 mm. below; retreat from tip Lower 24 by 9.5 \ greatest in lower spot, no notch. 11. Upper 24 by 11 /Gray tip; 4 mm. above, 6.5 mm. below (retreats as in 31 by 11 \ No. 12). Just a slight intersprinkling of black; terminal bar has a shade of a beginning. 10. 2b hv 13 Gray tip, 12.5 mm. 9.' Terminal 8 Whole width and slightly over shaft to lower web. Lower 24 by 13 Gray tip, 16 mm. Terminal 15 Whole width. 8.- 26 by 13 Gray tip, 17 mm. terminal encroaches, advancing from shaft 4 mm. H Terminal 18 [Whole width plus an advance on lower web, where only 19 by 14 1 a central triangular remnant of gray is left. 1 Full width of both webs. 6.' Terminal 19 Lower 15 by 13 \ A freckled patch blending with terminal, but differing in hue. 5.' 4. • 3. tj Terminal 19 Whole width, equal on both webs. 13 23.5 by 12 A freckled blotch; a little less dense than No 6. Whole width- Terminal No lower spot. All with terminal 23.5 Terminal increases downward; subterminal increases upward, to this plate in Chapter IV. — Ed.) (A fuller textual reference 16 ORTHOGENETIC EVOLUTION IN PIGEONS. ent stages of new characters. Mutation would be equally helpless, and each step would leave a gulf of discontinuity — something that nature seems to abhor. Turning from theoretical impasses to the facts, let us compare the two patterns. In the chequered pattern all the feathers are marked alike — no regional differ- entiation. In the other type we have a conspicuous local differentiation (pi. 2 and text-figs. 1 and 2) , suggesting at once a higher stage of evolution. 5 Chequered wings are to be found which vary all the way between a uniform marking and the barred type. If we arrange a number of unequally chequered wings in a series, running from the most to the least chequered, we shall see (pi. 3) that the pattern ap- proaches more and more nearly to that of two bars, as the chequers diminish in size and number. We shall notice that the pigment is reduced more rapidly in the anterior than in the posterior part of the wing. As chequers are reduced, they gradually lose their sharp ends and approximate the square or rounded form seen in the elements of the typical bars. The series shows a flowing gradation, that may be read forward or backward with equal facility. Darwin's view takes the bars as the starting-point and reads forward. Taking the chequered condition as the point of departure, the variation runs just as smoothly in the opposite direction. We here meet an ambiguity that is every- where present in color-pattern problems — an ambiguity that is frequently over- looked with disastrous consequences. The only way to eliminate the difficulty is to take our evidence from several different sources, and when agreement is found for one direction and disagreement for the other, the way is clear. As an experiment, we may take one or more pairs of pure-bred, typically barred pigeons and keep them isolated from chequered birds for several years, in order to see if the young ever advance toward the chequered type. Another experiment should be tried for the purpose of seeing what can be done by working in just the opposite direction. In this case we take chequered birds, selecting in each generation birds with fewer and smaller chequers, and rejecting the others, in order to see if the process of reduction can be carried to the condition of three, two, and one bar, and finally to complete obliteration of both chequers and bars, leaving the wing a tabula rasa of uniform gray color. If these experiments are continued sufficiently far, it will be found from the second experiment that a gradual reduction of pigment to the extreme conditions named can be comparatively easily effected, and that the direction of reduction will always be the same, from before backward; while, from the first experiment, it will be seen that it is hopeless to try to advance in the opposite direction, from the bars forward to the chequered condition. No variations will appear in that direction, but such as do appear will take the opposite direction, tending to diminish the width of the bars and to weaken their color. It is in this way that we must account for the existence of some fancy breeds in which the bars have been wholly obliterated. The direction of evolution can never be reversed. I have tried both experiments for eight years, and as both tell the same story as to the direction of variation, I am satisfied that further experiments will not essentially modify the results. Plates 4 and 5 indicate some of the results 6 in the reduction of the chequers. 'This situation is dealt with more fully in the explanations of plates 2 and 3. — Ed. 6 Most of the available illustrations for these results exist only in a series of small lantern-slides. The breeding data for the birds illustrated in plates I anil 5 are given in table 70, Vol. 11 ; some further breeding data are also given in Chapter IX, and another illustration in plate 16 of Vol. II. — Ed. PLATE 3 SO c3 oo 3 -3 2 S S ? o £^ a •2 a z * B ° -C bo a a *•£ o *_ IE O CD n a O 03 a a .2 ""' Si O a s . o r d «o ■O * OS en ^CO ° Jt? ^ so "3d r ■§ p ~ § § I * kK _£/ o* i> ■— ! *> .2 3 ,a *cc fc o THE PROBLEM OF THE ORIGIN OF SPECIES. 17 Text-figure 1. — Adult barred rock-pigeon (Columba livid), male No. 0. x 0.75. Hayashi del., Oct. 1904. Description of Bars and Size of Spots, October 1904. Second bar. First feather Second Third Fourth Fifth Sixth Seventh. . . . Eighth Ninth Tenth Eleventh. . . Twelfth Thirteenth . Fourteenth . Fifteenth. . . Sixteenth. . . Exposed (spot). Total. Upper= 8 mm. 10 mm. Lo\ver= 8 10 Upper=ll 14 Lower=10 14 Upper = 12 13 Lo\ver= 9 13 No upper, faintest dots. 12 12 14 14 14 12 Covered. 3 3 1 4 4 4.5 6 7 7 6 Distance to tip. 4 5 7 10 10 14 17 20 22 22 22 23 25 83 25 26 Remarks. Not shown in drawing. Spot is round and flecky. Scattered flecks only. Spot so thin as to be barely visible. Spot; single fleck. No spot. No spot. No spot. First bar. Terlials: First feather . Second Third Fourth Fifth Sixth Secondaries: Seventh Eighth to the Fourteenth . . Total length of spot. Upper=21 mm Lower=23 Upper=25 Lower=26 30 25 Upper=13 Lower=17 Exposed. 11 mm. 15 1 16 3 13 2 13 Covered. 10 8 24 10 27 12 11 4 Distance to tip. 2.5 mm 2.5 3 4 4 15 5 20 Remarks. No upper; terminal; begins and expands over shaft at tip a little Lower=20 Term. =18 Lower=18 15 Spot=thin flecks No spot. 11 10 15 20 23 24 | Term. above=ca. 12 mm Lower term. = 8 Term.upper=24 Term.lower=15 (As I do not find 14, and as there seem to be no missing secondaries, I think that the first upper tertial is the missing feather. No. 7 can be reckoned either to tertials or to secondaries. 18 ORTHOGENETIC EVOLUTION IN PIGEONS. Text-figure 2.— Adult male barred rock (C. lima), No. 7. x 0.75. Hayashi del., Oct. 28, 1904. bars have greatest width above and grow narrow downward (outward). Description of Bars and Spots, October 1904. In this bird the Second bar. Length, exposed. Total. Covered. Distance to tip. Remarks. First feather. Upper=15 mm. 17^ mm. 2^ mm. 1^ mm. Lower=15 17i 2h 3 Upper=16 18^ ■2i 4 Lower=13 m 4| 6 Third Upper=13^ 16 2 3 8 Lower= 1 1 15 4 10J Upper=12 14 2 11 Upper spot triangular and readies from Lower= 9 Yi\ 4-i 15.; shaft about two-thirds width of vein. Fifth No upper. Lower= 8 13J- 5i 19 Sixth Lower= 6 12 6 22 Lower= 6.5 14 ~\ 20 Eighth Lower= 5.5 11 f>\ 21 Ninth Lower= 4 10 6 22\ Spot begins to be fiecky. Tenth Lower= 2 8 6 22 Thin, fiecky. Eleventh .... Lower= 5 8* 8 20 Thin, fiecky. Twelfth Lower= 7 7 22 Thinner and fiecky. Thirteenth . . Few pale flecks 14th to 16th . First bar. Total length. Exposed. Covered. Distance to tip. Remarks. First feather. Upper=ll 2 9 2 Lower=16 3 13 .'. Second Upper=21 12 9 *i Lower=27 15 12 2 Upper=25.5 3 ijaj ..i -2 Lower=29 18 11 21 TJpper=30 a 25^ ^3 Lower=37 23 14 4 Fifth Upper=17 3 14 3, Lower=39 29 10 3 J 15 At shaft (and) At oulcr edge of web. Sixth Upper nearly gone Lower=24 10 14 18 Seventh .... Lower=16 si Ml 18 Eighth Lower=16 H "1 20 Thinner, but present, though not drawn as black, and limits not seen in figure. = 14 3 11 29 Thinner still; scattered flecks; terminal began on eighth, but first becomes plain here on ninth term. Term.— about 15 mm. 10th to 13th. No spot Term. =20 mm. I do not find more than 13 feathers. The last 6 are secondaries (8 13). The others (1-7) are tertials. I have examined this bird once before, but could find only 13 feathers. 1 think now that this is the true number, as No. (0) ( = fig. 2, this figure. — Ed.) agrees in this respect. THE PROBLEM OF THE ORIGIN OF SPECIES. 19 With reduction traveling from before backward, in the manner described, we get the bars in their typical number, form, and position, as one of the necessary stages of the process, and without appealing to de novo origin, incipient rudiments, etc. But if bars originated in such simple fashion — the direction of evolution being precisely the same as that of embryological development — if the theory of rudiments must be abandoned in this case, do we not meet the same theory again in any attempt to account for the chequers? What kind of rudiments could be imagined? We might assume that minute flecks of pigment first appeared, one in each feather; and then, further, imagine that these purely chance originations Text-figure 3. — Oriental or Japanese turtle-dove, Turtur oriental-is. x 0.5. Todadel., 1903. The color-pattern in black and white. The light-brownish or bronzed edges of the feathers given in white, the dark centers in black. This pattern is preserved in the feathers of the neck-mark. happened to have some slight utility, and that natural selection did the rest. But it is just as difficult to account for a small as a large origin de novo, and the smaller it is the more unfortunate it is for the theory of natural selection. If we seek refuge in the doctrine of mutation, are we better off? Mutation hides itself in the undiscoverable premutation, and so we have all the difficulty of an incipient stage, and no means of advancing by ordinary variation. Fortunately we are not driven to either alternative, for the chequers arise neither as mutations nor as rudiments, but by direct and gradual modifications of an earlier ancestral mark, which came with the birth of the pigeon phylum as a heritage from still more distant avian ancestors. 20 ORTHOGENETIC EVOLUTION IN PIGEONS. EXPIANATION OF PLATE 4 A. Male homer, No. 4, stray, 1900. Natural size Hayashi del., Nov. 1904. The male of a pair reserved to see if I could reduce chequered pattern to 3 bars. Result: could reduce to 2 bars, or 2 bars plus a vestigial third; that is, reduction was even for the whole wing. The 2-barred young (fre- quently obtained) bred true — gave 2-barred offspring. Thus far (Dec. 26, 1904, 5 years) there have been 5 young with only 2 bars and pale gray; 5 young nearly like male. The majority were intermediate to the male and female parents. The black (of drawings) is of course misleading, for bars vary all the way between almost black and dark gray . These many shades are ignored in ink drawings. But the size is given, and thin edges and very thin or pale spots by dots. Seen at a little distance, this male appears to have only 2 dark bars. Within a few feet we notice a few obscure and small vestigial chequers in the third bar and a few scattered chequers of about the same thinness in the scapulars. Spots a little stronger behind and above; weaker before and below. The bars are dark gray, not so dark as in the female mate. The gray ground of this male is considerably lighter than in the female. That is, with reduction of chequers comes a paler gray ground as well as less dark (paler) spots. Measurements of first bar {Dec. 28, 1904)- Feather No. Length exposed. Total length. Length covered. Remarks. mm. mm . mm. 1 7 18 11 2 18 24 11 3 17 25 8 4 20 27 7 5 18 21 3 6 15 19 4 7 16 18 2 Rather thin and freckled, except mid-third. 8 13 17 4 Mostly thin freckled, except in lower half. 9 6 12 6 Wholly and very sparsely sprinkled with pale. 10 Has only 2 or 3 obscure dots. From here on the tips of secondaries are darkened. Tips of upper web of sixth and seventh secondaries darkened ; the dark tip of eight secondary includes a little of lower web; this goes on more and more for lower web of rest of secondaries. Measurements of second bar. T Feather No. . otal igth. Length exposed. Length covered. Remarks. i nm. m in . mm. 1 10 9 1 Lower web; upper web few freckles. 2 17 16 1 Lower web; upper web about same as lower. 3 17 16 1 Lower web; upper web a little weaker than lower. 4 18 15 3 Lower web; upper web total length = 15 mm. 12 Lower shaft ; upper web only few 5 18 10 12 3 freckles; pale. Margin. Lower shaft; upper web no 6 19 10 12 2.5 freckles. Margin. Lower shaft; upper web no 7 18 11 2.5 freckles. Margin. 8 15 12 3 Flecked at both ends a little more than seventh feather. 9 13 10 3 Flecky throughout. 10 8 6 2 All thin-flecked. 11 Only a few thin, pale flecks. 12 Fewer and paler. 13 Vanishing. Third bar. Wholly vestigial and visible as pale dots or flecks only on 5 or 6 upper feathers. PLATE 4 A. Male homer, No. H4, stray, 1900. Natural size. Hayashi del., Nov. 1904. B. Female homer, No. K2, mate of No. H4 above. Natural size. Hayashi del., Nov. 1904. THE PROBLEM OF THE ORIGIN OF SPECIES. 21 Explanation of Plate 4 — Continued. B. Female homer, No. K2, mate of No. H4 above. Natural size. Hayashi del., Nov. 1904. This female raised from a male bought February 1897. He was fairly thickly chequered, but there was some reduction manifest in the field of lesser coverts. The mother was a stray ( 1898 1, strongly and quite evenly chequered. In K2 the scapulars are all double-chequered, but the spots are considerably reduced, paler in front than behind and more fiecky-edged. The spots are fairly long behind, but the points are more or less rounded. There is a pretty well defined row of spots in the third bar, and a few spots of vestigial dimensions and paleness are visible in the fourth row of feathers. On lifting the feathers I find many spots that are wholly concealed. The rest of the small coverts are mostly without visible spots. In the male the second bar is cut so straight that no points are seen, while in the female many of these spots are pointed, so as to give a serrate edge, particularly near the middle. Measurements of first bar (Dec. 28, 1904). Feather No. Total Length. Length exposed. Length covered. Remarks. mm. 77i m. mm. 1 16 6 10 2 26 15 11 3 31 19 12 4 35 20 15 5 38 25 13 Has only a little black on upper web at tip. 6 35 25 13 Begins to be black on lower tip, and has more black on upper tip. 7 36 25 11 Black of tip increased in both webs. [38 27 11 Black of tip increased on upper 8 web and on lower web it co- alesces with spots. [32 17 Original spot still distinguishable, although tip is all black. Nos. 38 and 27 apply to only the proper bar-spot, which is sub- terminal, as in the tertials 9 45 15 10 45 15 11 45 15 12 45 15 Secondaries. 13 43 15 14 40 15 Measurements c / second bar. Feather No. Total length. Length exposed. Length covered. Feather No. Total length. Length exposed. Length covered. 77! 77J . 77(771. 777771. mm. 777777. mm. jl 19 13 6 9 23 18 5 2 20 14 6 10 22 16 5 3 22 17 5 11 21 15 6 4 25 20 5 12 21 14 7 5 27 20 7 13 21 13 8 6 27 20 7 14 20 11 9 7 26 20 6 15 20 11 9 8 25 19 6 16 15 10 5 The spot of the sixteenth feather is thin, but still quite a spot except for wide, flecky margins. Gradually, as we pass upward, the spots become darker and fuller (i.e., the flecky parts are less in extent), until at the ninth feather we have a well-developed spot which is pointed but ragged-edged distally and proximally. Third bar. Most of spots have 6 to 8 mm. uncovered. Total length, 10 to 12 mm. In central part of field, where no spots are to be Been, there are many spots entirely concealed which are only seen when feathers overlapping are lifted. 22 ORTHOGENETIC EVOLUTION IN PIGEONS. Explanation of Platk 5. A. Female homer, No. 3, hatched Mar. 31, 1901. x 0.8. Hayashi del., Nov. 1904. Parents of the three birds of this plate are shown in plate 4. Wing from a large bird of fine form; very pale; primaries inclining to whitish. The bars in this bird are obscurely subdivided by lines (alternating light and dark narrow bands) or bands that are not perfectly transverse, but inclined so that those of two sides meet at a very obtuse angle (i.e., they are what were later called "fundamental bars." — Ed.). Third bar. — First, second, and third feathers each have two spots — both weak and pale, and plainly obsolescent (ragged edges and light lines crossing them). Lower down only one spot to the feather (lower web), pale and increas- ingly thin and roughly outlined. Elsewhere in front of this bar the color is pure — very pale gray; no spots to be seen. A few vestiges found on lifting feathers. Scapulars: In front no spots; in middle spots are pale reduced vestiges; behind they become more distinct, but they are everywhere weak and without clear outlines. Second bar. — On right side some of middle spots are pointed, but on left side corresponding feathers show spots with points disappearing. Feather 2. Length of spot, 13, exposed; total length, 20 mm. 3 and 4. Length of spot, 13, exposed. 5. Length of spot, 11, exposed; point obsolescent. Length of spot, 21, exposed; point intact but distal edge flecky. Length of spot, 12, exposed; point cut off as an island. About same, but island thinner. About same, but thinner. Shorter and still thinner. Length 8 mm., pale and flecky throughout. 12 to 15. Thinner and vanishing on last. About 4 mm. of the mid-spots are covered. Measurements of first bar. 6. 7. 8. 9. 10. 11. Feather No. Total length. Length exposed. Length covered. Remarks. mm. mm. mm. 1 23 16 7 Lower web; spot on upper web about half as long, terminal reduction from proximal side. 2 27 21 6 Upper web; spot; terminal 16 mm. 3 30 19 11 Upper web; spot; terminal 20 mm. 4 30 20 16 Upper web; spot; terminal 18 mm., very thin. 5 31 20 11 Upper web ; no spot. 6 31 20 11 Upper web; no spot, but faint shadow of a tip spot; spreads over onto lower web. 7 32 20 12 Terminal spot enlarges and creeps over to lower web more ; length about 1.4 mm. 8 24 13 11 Whole tip (ca. 19 mm.) now dark, but original spot is darker, so its limits are plain; it is re- treating from tip, being now 19 mm. against 14 mm. in seventh feather. 9 15 12 3 Whole tip (ca. 24 mm.) dark; original spot is flecky through- out. 10 14 10 4 Whole tip dark; original spot very pale flecky, almost gone. 11 Just a trace left; tip dark, 25 mm. 12 No traces; tip dark, 25 mm. 13 Tip dark, 24 mm. 14 Tip dark, 23 mm., about one-half as dark as preceding feathers. Eighth spot is decidedly paler than in seventh feather. This sudden drop to pale spots at eighth seen again in figure U of plate 4. This bird has made a decided advance over the dam, but she is not quite so far along as the sire. The second and third bars are narrower than in the dam. The first bar is wider than in the sire, but the second is about the same. PLATE 5 A. Female homer, No. 3, hatched Mar. 31,1901. X 0.8. Hayashi del., Nov. 1904. Parents of the three birds of this plate are shown in pi. 4. B. Male homer, No. 7, hatched May 1, 1904. X 0.8. Hayashi del., Oct. 1904. C. Female homer, No. 8, hatched May 1, 1904. X 0.8. Hayaslii del., Oct. 1904. AJUiwICtt P^n r* THE PROBLEM OF THE ORIGIN OF SPECIES. 23 This ancestral mark is a dark spot filling the whole central part of the feather, leaving only a narrow distal edge of a lighter color. This mark is still well preserved in some of the old-world turtle-doves — best in the Oriental turtle-dove of China and Japan. The chequer of Columba livia differs from the dark center of Turtur orien- talis (text-fig. 3) only in form and in having a lateral position. Typically this spot appears in pairs, one on each side of the feather. The two spots represent the two halves of the old central spot, which becomes more or less deeply divided by the disappearance of pigment along the shaft of the feather. This change begins at the tip of the feather and advances inward, but usually more rapidly along the shaft than at the sides, thus resulting in two chequers with more or less pointed tips. The direction of change again coincides with that of embryonic develop- ment — the tips of the feather, where it begins, being first in order of development. In many chequered rock-pigeons we may find in the dorsal (inner) feathers of the bars undivided central spots, which pass gradually into the typical chequers as we pass towards the lower (outer) ends of the bars. Transitional stages of Explanation of Plate 5 — Continued. B. Male homer, No. 7, hatched May 1, 1904. X 0.8. Hayashi del., Oct. 1904. The male bearing this wing resembles his sire in size and general form. The chequers on the scapulars are stronger and more numerous than in the sire, but fall far below the dam. The third bar has about the same number of spots, but these are here a little stronger. The second bar inclines to red (red = light shading; black or rather dark gray = black). The shaft-line and the lower edge are also reddish gray in three or four of the upper (inner) tertials and long coverts. Measurements. Feather No. Second bar. First bar. Length exposed. Remarks. Length exposed. Remarks. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 mm. 13 13 14 15 17 21 14 16 14 10 8 8 7 6 5 mm. 15 20 25 24 19 17 17 15 11 (?) (?) (?) (?) 2 spots. 2 spots. 2 spots. 1 spot; no upper spot. Spot on upper tip begins. Spot at tip spreads over to lower. Reduced to red-gray. Do. Do. See below. Flecky. More flecky and thinner. Thinner. Thinner. Thinner. Few flecks. The eighth to fourteenth secondaries of the first bar have spots almost too indistinct to measure with certainty. (See preceding figure.) 1904. The quills are whitish, but C. Female homer, No. 8, hatched May 1, 1904. X 0.8. Hayashi del., Oct This female is pale, though not quite so pale as homer No. 3 (fig. A, this plate), less so than in homer No. 3. Third bar is here scarcely noticeable. Second bar begins with 11 mm. width, increasing in second feather to 12 mm. and in third to 13 mm. It runs on nearly evenly to the seventh feather and then becomes gradually thinner. First bar has its greatest width on fourth and fifth feathers— 22 and 20 mm. respectively. The tip spot begins on both upper and lower webs of the sixth feather. (The details of the breeding history of the three birds shown on this plate arc given in table 7b .if Vol. II. — Ed.) 3 24 ORTHOGENETIC EVOLUTION IN PIGEONS. various degrees thus connect the derived with the ancestral type in one and the same individual, and so demonstrate that the two specific marks are not separated by impassable mutation-gaps. While it is not necessary to go beyond the wild rock-pigeons and the multitude of domestic races descended from them to learn that nature has here pursued one chief direction of color variation, always leaving an open door, however, to minor modifications and improvements through natural and artificial selection, it is nevertheless highly instructive to make a comparative study of the whole group of wild pigeons, in both adult and juvenal stages. It is in this field that we find the same lessons amplified and repeated in multitudinous mm x Text-figure 4. — Adult female Florida ground-dove. Chamoepelia passerina. Natural size. Hayashi del., Apr. 1903. The feathers of head and neck are edged with pale dark edges, crescents not so strong as in Scardafella inca. Note the dark centers in feathers of breast and foreneck and the reduced spots of the wing. For fuller description of these feathers see pi. 28. In the anterior part of the wing — in the region of the oblique liar of some related forms (see under Peristerinae, Chapter V. — Ed.) — there is a noticeable persistence of spots. ways, confirmation confirmed, convergence of testimony complete. It will be suf- ficient here to cite a few examples. In the little ground-doves (Chamoepelia passerina) of Florida, Arizona, Cali- fornia, Central and South America, and the West Indies, we find the turtle-dove pattern preserved in the whole breast region (text-fig. 4) and in the anterior, smaller coverts of the wings, while in the posterior portion of the wings we meet with lateral spots or chequers, of higher finish than in the rock-pigeons. In many coverts of the wing we find the dark centers more or less reduced, with the distal ends of their remnants in various stages of conversion into lateral spots. Here again we find striking proof of gradual change from one specific type to another. THE PROBLEM OF THE ORIGIN OF SPECIES. 25 In the brilliant bronze-winged pigeon (Phaps chalcoptera) of Australia we have still another combination type, in which iridescent chequers coexist with the original dark centers (pi. 6, fig. A). Here the chequer seems to arise by direct differentiation of a lateral portion of the dark center (pi. 7, fig. B), the latter still occupying the original field and forming the ground within which the chequer appears as a more highly colored spot. While the dark center does not suffer any reduction in its field, it does lose considerably in intensity of color. The metallic spots are there- fore probably built up by concentration of pigment at the expense of the dark Text-figure 5. — Chequers of wing of Juvenal male passenger-pigeon, Eclopistes migratorius. x 0.! Hayashi del., Apr. 1S97. Text-figure 6. — Chequers of wing of adult female passenger-pigeon, x O.S. Hayashi del., Feb. 1897. centers. As these birds make great display of their colors in the breeding-season, 7 this departure from the orthogenetic trend of development may be attributed to natural selection. 8 The wild passenger-pigeon (Edopistes) bears chequers closely resembling those of the chequered rock-pigeon, in form, color, and distribution. In this species the sexes are distinctly differentiated in color (see pis. 28 and 29, Vol. II) ; and we have 7 This topic is treated in Vol. III. — Ed. 8 Soon after the above was written the author received and studied (and had illustrations made of) specimens of another bronze-wing (Phaps elegans), which bears a relation to the much-chequered 6pecies chalcoptera, quite similar to that borne by the two-barred rock-pigeon to the chequered rock. The figure for Phaps elegans is therefore also reproduced in plate 6, figure B. — Ed. 26 ORTHOGENETIC EVOLUTION IN PIGEONS. for comparison three stages in an ascending series, namely, the juvenal (text-fig. 5), the adult female (text-fig. 6), and the adult male (text-fig. 7). As in so many other birds, the male makes the widest departure from original conditions; the female occupies a lower plane; the young are nearly alike in both sexes and may be said to recapitulate ancestral conditions with less modification than is seen in the adult of either sex. Explanation of Plate R. A. Left wing of bronze-winged pigeon, Phaps chalcoptera ; probably female No. S 1, red. Hayaslii del., May 1899. x 0.75. Description, December 25, 1S98. First bar. — Some of the feathers are imperfect. This bar has four brilliant spots. There is a finish and definition here that excel anything I have seen thus far on the pigeons. Feathers Nos. 14 and 13 have no spots. No. 12 is darker in lower web, but no distinct spot. Nos. 11 to 6, six feathers, which in closed wing are seen in upper part, are thus marked. Nos. 11 to 8 have the four most brilliant spots. These spots have a rich purplish reflection, in some lights show- ing only black. The inner secondaries have in addition to the spots a steel-blue reflection on the outer web. In all these spots the fawn-color of the lower edge and tips of the feathers extends from the edge up into the web, forming a beautiful border on the terminal end of the spot, half converting the spot into an ocellate form. Notice that all these spots are more or less pointed posteriorly. No. 11 spot is ovate in shape, with small end behind. It is bordered with a velvety black just within the light- fawn border behind. No. 10 has a boot-shape with toe pointing backward; longest measure, 23 mm.; transverse, S mm. No. 8 spot similar in shape, but more pointed at toe. No. 6 spot is oval oblong and duller. These four brilliant spots are placed where they can be seen to best advantage in the closed wing. Second bar. — There are 6 brilliant spots. Here the lower 5 feathers have a metallic green reflection on the lower web, which becomes stronger upward, giving a spot on 5 that has an oval center of a bronze-green reflection. This becomes stronger in Nos. 6 and 7, and in No. 8 enlarges to an oblong form 10 by 11 mm., with a light-fawn streak as border behind. In No. 9 we have the largest spot — squarish, measuring 16 x 11 mm. From here on to No. 13 the spots become smaller. From No. 8 to No. 12 the spots show more or less obtuse points backward. The spot on 13 is not pointed; it is the smallest, measuring 7 by 8 mm. These spots, like those on the median and lesser coverts, are on a golden bronze-green, often reflecting a rich coppery red. Here, as before, the spots are developed on the upper middle feathers, where they can be seen to host advantage . Third bar. — Has 14 brilliant spots. This is the most highly developed bar, so far as number and brilliancy go. The 14 brilliant spots are all bluntly pointed behind. They increase in size and color from below up to the eighth (the largest) and then diminish gradually to the last or fourteenth. Here again it is the middle feathers that are richest. Sizes: Lower spot (1) 12 by 6 to 12 mm., largest (8) 10 by 9 mm., smallest (14) 2 by 3 to 12. The rounded posterior ends encircled with velvet-black and light fawn-color, giving the spots quite an eye-like appearance. In Nos. 8 and 9 the spot extends a little above the shaft, reaching highest point in No. 11, where it has the bronze-green reflection. This seems to be a rudiment of an upper spot. Fourth bar. — Has 11 brilliant spots, beginning in No. 3 (perhaps on No. 2), becoming largest on Nos. 5, 0, 7, and 8, then diminishing upward to a small circular spot of 3 by 4 mm. on No. 13. The spots are not quite so many nor quite so large on the average as in third bar. Most of them are bluntly pointed backward. These spots begin to cross above shaft, slightly in Nos. .5, 6, 7, and more plainly in Nos. 8 and 9, and in No. 10 1 notice the same bronze reflection in the small upper spot. This spot becomes black in No. 1 1 , and is not seen above. I find that the number of spots in a series, or row, differs in the two individuals I have. The specimen described has smaller spots than the other, and they are not so rich in color. B. The bronze-wing, Phaps elegans. Adult male, x 0.75. Hayashi del., Sept. 1904. In first bar 3 spots. Note and description, Sept. 11, 1905. I have two males that have just died. In one of these I find no distinct chequers in the tertials on the left side, but on the right side there are two obsolete chequers. The second and third bars are about as well developed as in specimen drawn. In the other I find two very obscure spots in the tertials of the left side and only one such spot on the right side. I find no spots in the second bar of the right side (some were lost on left sidl — presumably they were about the same). Thus while the first bar is almost obsolete, the second bar has two more spots than shown in the figure drawn. PLATE 6 . SSfe A. Left wing of bronze-winged pigeon, Phaps chalcoptera; probably female No. SI, red. Hayashi del., May 1899. B. The bronze-wing, Phaps elegans. Adult male. X 0.75. Hayashi del., Sept. 1904. bar 3 spots. X 0.75. In first AMofnltapJiWEOrr THE PROBLEM OF THE ORIGIN OF SPECIES. 27 Text-figure 7. — Cheq- uers of adult male pass- enger- pigeon. Tod a del., Nov. 1903. x0.5. In birds taken at random, I count in the left wing and scapulars 90 chequers on a juvenal, 51 in an adult female, and 25 in an adult male. This is pretty conclusive evidence that chequers are, or have been, disappearing in the species. Not only the number but also the size of the chequers has been reduced. In the female the chequers are for the most part two or more times as large as in the male. The reduction in both respects has been greater in the anterior than in the posterior half of the wing, and greater along the lower edge than in the middle and back regions. In this species we may recognize at first sight the homologues of the rock-pigeon bars. On the secondaries of the female we find the homologue of the posterior bar, and on the first row of long coverts the homologue of the anterior bar. The latter is scarcely recogniz- able as a bar, for we see only 5 or 6 chequers in the upper half of the row, the lower half being without chequers. Nevertheless, this row represents, so far as it goes, the elements of a bar which is already too far gone to have even a chance to attain the finish of a per- fect bar. 9 On the secondaries the chequers fall into juxtaposition, forming a continuous bar, with an irregular posterior outline, which indicates that the chequers have been unevenly re- duced from behind. It is a rudely finished bar which has sunk below the horizon of utility, if it was ever above it, and is now facing ulti- mate eff acement. The reduction has advanced further in the male, with no improvement towards regularity of outline. Here it becomes quite certain that effacement advances from all sides, leaving but a small remnant of a bar confined to two or three feathers. Glancing at the wing as a whole, in both young and old, it is plain that the process of obliteration is in progress over the entire chequered area. The elongated, sharp-pointed marks of the earlier pattern have rounded tips in the adult; the posterior bar is roughly emarginated; the number of chequers is reduced by half or more; and some of the remaining ones are but little more than mere dots. It is also equally manifest that the process of reduction is making more rapid progress in the fore part of the wing and along its lower edge than elsewhere. There can be no mistake here as to the direction in which the phenomena are to be read. '' In the young, the chequers of this row are more numerous and much more sharply pointed at the ends. In both respects the juvenal pattern approaches more nearly a condition of general uniformity. 28 ORTHOGENETIC EVOLUTION IN PIGEONS. The direction is as certain as that the adult male stands in advance of the adult female, and still more in advance of the young bird. The significance of the case lies mainly in the fact that it is not an isolated or exceptional one. Many other species tell more or less perfectly the same story. A parallel case, only carried still farther in the same direction, is found in the mourning-dove (Zenaidura carolinensis). The adult male and female differ but slightly, each having only about a dozen chequers visible on each side. These are confined to the scapulars and to a few feathers at the posterior upper edge of the wing. In the young they are more numerous (pi. 7, fig. A), but less so than in the young passenger-pigeon. The middle and fore parts of the wing in the adult have no visible chequers, but a few concealed ones which may be seen on lifting the overlying feathers. These concealed chequers, and other differences between old and young, show that the species has its origin in a chequered stock, and that its history has been analogous to that of the passenger-pigeon. The white-faced pigeon (Melopelia leucoptera) is a most instructive form. Although much more highly accomplished than the mourning-dove in the arts of display of form, feathers, and voice, it has suffered a complete effacement of the chequers it once possessed in common with other members of the family (see text- figs. 8 and 14). indubitable proof of this is to be seen in the juvenal feathers, which in some cases exhibit a few pale vestigial spots in the last two rows of long coverts, at points where the chequers are usually best developed in chequered species. Another striking proof is to be found in the coverts and scapulars of the adult bird, where we find, on lifting the feathers, distinctly outlined areas, corresponding in shape and position with reduced chequers, but from which the black pigment has disappeared. These vestigial outlines, structurally defined, were first noticed in a female bird of a dark shade. 10 The outlines were more perfect than in lighter birds obtained from Arizona and California. 10 Captured in Jamaica by Dr. Humphreys. Explanation ok Plate 7. A. Wing of juvena! mourning-dove, Zenaidura carolinensis; age 4 weeks. Natural pize. Havashi del., June 1897. Left wing in first plumage. The tertials have 2 black spots, the dorsal one having no sharp bounding-line. The darker spots are at the upper part of the posterior bar. The next bar has 8 spots, and next 8. Most of the scapulars have two spots, and many of the anterior lesser coverts bear indistinct spots. Spots grow darker backward. These spots are strikingly like those of the young Ectopistes. Feathers edged with a dull buff -color. The quills are much like those of Ectopistes, but the edging is not of so rich and deep a color. B. Wing of juvenal bronze-wing, female, Phaps chalroptera. Natural size. Hayashi del., Dec. 1005. Left wing at four weeks. (1) Oblique streak which has about 10 feathers in one or two rows. These lie just in front of the fourth row in wing-coverts, counting from behind. The feathers of the streak have each one black spot on lower web, only slightly iridescent, and they are already long enough to reach over the tips of some feathers of the fourth row. (2) On the tertials it can be seen plainly how the spots arise in or from the dark center of the turtle-dove pattern, a pale transverse band, running from the pale edge of the feather inward for all lengths, from a part of a millimeter to one-half or two-thirds of the width of the outer web. The cross-band is of same color as edge of feather. It means that pigment fails to appear here, and, as if in compensation, more pigment is laid down in front of the band, thus producing the spot or element of the bar; and this spot in adidt feathers is built much more strongly — more black pigment and strong iridescence. In another specimen of about the same age, also a female, I measured these cross-bands and found them to be from 2 to 3 mm. in width and 1 to 8 mm. in length. They stand IS mm. distant from the tip of the feather. In the specimen drawn 1 see these bands in the row of long coverts, but not so well defined. Evidently these cross-bands or bars have had much to do in shaping spots and in giving eye-like borders (diamond-dove, adult bronze-wing, etc.). The wing drawn was injured a little at outer edge, PLATE 7 1 n * A A. Wing of juvenal mournings-dove, Zenaidura cardinensk ; age 1 weeks. Natural sue. Hayashi del., June 1897. B. Wing of juvenal bronze-wing female, Phaps chalcoptera. Natural size. Hayashi del., Dec. L905. B idmtn THE PROBLEM OF THE ORIGIN OF SPECIES. 29 Similar vestiges are present in the mourning-dove, and here their identification as marks formerly filled out with black pigment is freed from every shadow of doubt by chequers in all stages of obliteration. (The conditions presented by Zenaidura and Melopelia are fully discussed and illustrated in Chapter V. — Ed.) The large wood-pigeon (Cohmiba palumbus) of Europe has departed still more widely from the turtle-dove type, having lost all its black spots except a few in the neck patches, which have retreated so far from the tips of the feathers as to be concealed (see text-fig. 6, Vol. II) . The gray plumage and the white streak along Text-figure 8.— Adult white-winged pigeon, Melopelia leucoptera. x .77. The feathers at the lower outer edge of the wing are white. No chequers in the adult, but structural imprints of these in a few feathers of the wing. the edge of the wing mark a plane in the evolution of this bird very nearly identical with that of the white-winged pigeon. A little higher plane has been reached by our band-tailed pigeon (Cohmiba fasciata) of the Pacific coast, which is also a species of turtle-dove derivation, 11 as shown in the neck-markings (see text-fig. 5, Vol. II) and in the voice and behavior. These illustrations, which could be extended into the hundreds, may be con- cluded with two cases, representing wide extremes, yet governed by the same law of progressive orthogenetic variation. 11 Minute blotches of black were found in the longer scapulars of a few individuals. These are probably atavistic reminiscences of lost spots. 30 ORTHOGENETIC EVOLUTION IN PIGEONS. The crested pigeon of Australia (Ocyphaps lophotes) stands at the uppermost limit 12 in the number of bands and in the perfection of finish. There are 11 or at most 12 parallel bands crossing the wing and scapulars transversely, each bend marking a single row of feathers with the regularity of zebra stripes (pi. 8, fig. A). The width of these bands increases from before backward, beginning with a width of about 0.5 mm. and reaching 4 to 5 mm. on the tenth band. The eleventh band, located on the long coverts, is especially interesting, as it begins above with narrow elements, like the preceding, but is continued, from the third or fourth feather onward, by elongated chequer-like spots. This band or bar is the homologue of the anterior bar in the rock-pigeon, and furnishes a standing picture of transitional continuity from one character to another, at the same time settling beyond dispute the direction variation has pursued. So clear and decisive is the case that one might safely predict that this entire bar is destined to be reduced to the narrow-band type seen in the fore part of the wing. We have only to turn to a closely allied species, the white-breasted crested pigeon 13 (Lophophaps leucogaster) to find that it has already realized the prediction to the full, having every chequer in this row converted into a typical band-element (text-fig. 9). 12 The short statement on Ocyphaps given here is amplified by the whole of the following chapter. — Ed. 13 This bird is comparatively rare, and I have seen but a single pair that recently came to hand through the kind- ness of Mr. Frank M. Chapman. Explanation of Plate 8. A. Wing of adult male crested pigeon, Ocyphaps lophotes. Natural size. Hayashi del., Dec. 1898. This bird is nearly 2.5 years old. First bar. — Comprises 6 secondaries and 8 tertials. The elements of the bar are long, broad spots on lower webs ; iridescent; edged at tips with white. The iridescence vanishes on fourth feather. The lower web is about equally dark on Nos. 1 to -t, darker on No. 5 and slightly iridescent; iridescence brilliant on Nos. 6 to 10; only a little purple iridescence on No. 11. No spots on Nos. 12 to 14. Second bar. — Formed of 15 long coverts: (1) Black narrow bar on No. 14; is of same character as on median and lesser coverts. (2) Homologous with the broad iridescent spots of Nos. 13 to 5. (3) Upper spot or bar appears on No. 14 and disappears on No. 8. (4) Diminishes upward and downward, being strongest in mid-feathers, Nos. 12 to 8. Third bar. — Formed of 12 median coverts. (1) Bar consists of two parts, as seen on No. 12 and as shown in form on Nos. 10 to 2, and again on No. 1. (2) Bar widens in mid-feathers, but does not expand as in second bar; does expand more than in the fourth bar. (3) Bar of lower web on Nos. 10 to 2 is iridescent in some lights, but seen en /arc it appears deep black. Iri- descence then diminishes from behind forward, and does not extend to upper web in third ami first bars. It is inter- esting to find the iridescence of third bar invisible in most lights, and, when visible, sharply limited to lower web. The double nature of the bars is evident on nearly all the lesser and median coverts, but the lower half becomes more pronounced than the upper in the second bar, and in the first bar no upper spots are seen. Here we have remark- able variations in the spots. Fourth bar. — Formed of 11 feathers: (1) The bar consists of two parts, best seen on upper feathers (No. 11 ). (2) The lower half is iridescent, seen in favorable light. (3) The width of the bar is quite even and plainly narrower than third bar. (4) In form one can say with respect to these four bars that they become more perfect from behind forward; in color brilliancy they improve in the contrary direction. (5) In first bar, iridescence runs from No. 11 to No. 5, Nos. 10 to 7 being the larger and more brilliant. In second bar, iridescence runs from No. 13 to No. 5, Nos. 13 to 6 being the larger and more brilliant. In third bar each feather is marked on both webs, except the first, and the last imperfectly on upper web. In fourth bar each feather has a bar crossing both webs. The iridescence in the lower web is stronger in thud bar than in fourth. B. Wing of juvenal female crested pigeon, O. lophotes. Age 19 days. Natural size. Hayashi del., Aug. 1905. The ashen-gray ground-color contrasts strongly with the adult coloration (see above, fig. A). All feathers with pale or white tips; black pigment concentrated immediately in front of these. PLATE 8 A. Wing of adult male crested pigeon, Ocypha/px lophotes. Natural size. Hayashi del., Dec. 1898. B. Wing of juvenal female crested pigeon, 0. lophotes. Age 19 days. Natural size. Hayashi del., Aug. 1905. THE PROBLEM OF THE ORIGIN OF SPECIES. 31 Moreover, the transformation has already begun in the first feather of the next and last row, so that the same prediction could be extended to this bar, which is the homologue of the posterior bar in the rock-pigeon. Glancing again at Ocyphaps, and looking at the wing as a whole, the course of transformation, its mode, direction, and future termination are all very clearly defined. The wing-pattern, as shown especially in the light edges of the juvenal plumage (pi. 8, fig. B), takes us clear back to the turtle-dove type. Next came the chequered pattern (text-fig. 10) , similar to that of the primitive rock-pigeon. Reduc- tion of pigment, proceeding from before backward, fashioned the bilateral chequers from the unicentral spots. The reduction kept on in the same direction, shorten- Text-figure 9. — Adult white-breasted crested pigeon, Lophophaps leucogaxter. Natural size. Hayashi del., Oct. 1902. There are only four marks on tertials. The long tips of feathers are dotted for cinnamon color. The black bars are preceded by pale gray bars, not shaded. This bird stands in advance of the common crested pigeon, as shown by absence of elongated marks on the long coverts. ing the chequers and transforming the rows successively into narrow bands, eventu- ally reaching the eleventh row, where we find only one or two complete steps, followed by a graded series of four to six steps, less and less decided, until we lose every trace of them. So finely graded are these steps in some females that it is difficult to locate the vanishing-point. Unless the process of transformation is arrested by the extinction of the species, or through the intervention of some more potent modifying influences than have thus far appeared, the fate of both posterior bars is irrevocably scaled. Granting that natural selection may be credited with strengthening the iridescent splendor of these bars, I believe that the orthogenetic influences are bound to prevail here as in the white-breasted species. 32 ORTHOGENETIC EVOLUTION IN PIGEONS. Explanation of Plate 10. Left wing of aduit stock-dove, Columba anas, No. 2. x 0.9. Hayashi del., Dec. 1904. The bars of both wings shown in figure B. One of three adult stock-doves captured by Mr. Lane, of Cambridge, England. In the figures the first of the tertials has been omitted. It is now present in the bird (Mar. 12, 1005) This feather readies as far back as the longest scapulars, and hence to distal edge of the spot on the second feather, but it covers only the upper (inner) third of the spot. Measurements of first bar, left wing. Feather No. Total length. lCxpn.-i'd ,, , Distance Covered. , , .. to tip. Remarks. m m . mm. m m . 3 (No upper \ Lower 9 Upper third 1 Spots rhombic, with the lower (outer) \ distal and upper (middle) proximal 9 Lower an- 11 terior angle J acute. 4 15 14 1 nun 12 Upper, smaller, and weaker; although rhombic, angles are rounded and outline is not sharp. 5 10 10 linn 20 No upper spc its; upper edge of each spot is of course covered by the next feather above in all cases The first bar is very pale — about half as dark as second bar; it appears too black in ink. On the right side only two spots were found. Second bar. Feather No. 1, omitted in figure 1 (correct in fig. B). No. 2, shows mere trace of spot; would not be noticed. No. 3, spot middle of lower web, roundish, 9 mm. long, usually entirely covered; distance to tip, 13 mm. No. 4, spot about 14 mm. long, only 2 mm. exposed; distance to tip, 14 mm. No. 5, spot about 11 mm. diameter, only 2 mm. exposed; distance to tip, 18 mm. Measurements of third bar. Two minute rudiments, wholly covered; upper one occurs on fifth feather of row; distinct but not sharply out- lined, 5 mm. long by 2.5 mm. wide; 13 mm. distant from tip; lower one a very thin freckle of about same size as upper; would be overlooked if one were not guided by the upper spot. Explanation of Plate 11. Left wing and bars of both wings of adult stock-dove, Columba asnas. No. 3. x 0.9. Hayashi del., Nov. 1904. One of three adult birds from Mr. Lane. The first tertial is not shown in figure A; correctly shown in figure B. No upper spot in either bar. The distal side of all spots straight-slant in both bars; proximafend rounded in all spots of both bars. The first bar has three weak spots. Ink makes them too strong. They are so faded that at certain angles they almost vanish. In the fon rt h feather is found the strongest of the three spots. Measurement of left wing shown in plate 11 Feather No. Total length. Bxposed. Covered. Distance to tip. First bar: 3 mm. 12 17 14 9 11 13 13.5 9 mm. 9 10 11 3.5 5 6 6 4 m m . 3 7 3 5.5 6 7 7.5 5 mm. 10 15 19 9 13 15 15.5 19 4 5 Second bar (5 spots) ; 3 4 5 6 Third bar has four spots, all concealed; they lie on fourth, fifth, sixth, and seventh feathers; are rounded remnants, with distance from tip varying from 14 to 20 mm.; second and third spots strongest, first and last much weaker. PLATE 9 T -F u i. c — t - , r. - C 5 , >. X — ~ A - cd +-• :. > p GJ i<— i nc — o 3: i — ta X — ■3 - H -5 ~ ~ ti £ _ O ,~ X a js _z 4- — +-> r« O X c3 S-H bo r. "C 1 — 1 :- 1 — s- 71 X ~ ~ +j — - ^c ~ — c ~^ .— — i iC t£ ~ . C* .= -<— f ~ *'" ' .^ X r; § O +3 +j >, _^ - 1 — - ~ *" ?1 >: c **■ — ^ 7— 1 — - X T3 - O '? X* cs - FH X ■~ :/- od — _J^ C £ -4-3 c — .2 p. O Ph ^r — ^ 0^ c *o ■+j >. d - .5 2 X *= X — r; /- » X P = 2 •s T-t J: O o 73 - '- o o 5 H — PLATE 10 -# - r. r-* 6 0) - - . ^ a> o — -a H Js o o -t— X — 3 — cS S4-« o 60 P PLATE 11 B Left wing and bars of both wings of adult stock-dove, Columba arnas, No. 3. X 0.9. Hayashi del., Nov. 1904. One of the three adult birds from Mr. Lane. The first tertial is not shown in fig. A ; correctly shown in fig. B. No upper spot in either bar. The distal side of all spots straight-slant in both bars; proximal end rounded in all spots of both bars. The first bar has three weak spots. Ink makes them too strong. They are so faded that at certain angles they almost vanish. In the fourth feather is found the strongest of the three spots. WfimlCoIUttnwt THE PROBLEM OF THE ORIGIN OF SPECIES. 33 But is there any direct proof that the transformation is actually making progress to-day? May not these transitional steps go on appearing generation after genera- tion, without ever making any permanent progress? We have to concede that we can not follow the processes that reveal themselves in steps. We can at most only see what is done, not the doing. We are entirely in the dark as to the time required to carry the change through a single row of feathers. But we know that this has been done in three other species of the same family. We see after it is done, not before, that the transitional steps appear in the next and last row. Moreover — and this is as close as we can hope to get to actual seeing — we find that progress of just the kind we are looking for is certainly made in passing from the ju venal to the adult plumage. This is an ontogenetic change of a few weeks, which we can easily demonstrate by experiment to be pro- Text-figure 10. — Juvenal male crested pigeon, Ocyphnps lophntes. Age 15 days. Natural size. Hayashi del., Dec. 1898. This shows, in the first row of coverts, the transition between long spots (longitudinal streaks of black) and cross-bars (better than female juvenal in pi. 8. — Ed.). In this species the cross-bars are subterminal instead of terminal (as in Geopelia). Compare with adult Ocyphaps (pi. 3) and Lophophaps (text-figure 8). gressive and continuous. The corresponding phylogenetic advance has left no other record, and hence we only know that it took time — that it was not a momentary salt. In the adult plumage, one or two full steps are taken beyond the juvenal stage, and taken precisely at the points premarked by transitional steps. The number of transitional steps is increased at the same time. 14 As the next and last illustration, I take a case in which the bars are verging to complete obliteration. The well-known wild stock-dove (Columba oenas) of Europe may serve as a convenient and instructive example. In this pigeon we find that reduction of the chequers has swept over the whole bar, leaving nothing except a few obsolete spots, which we recognize as vanishing elements of bars formerly more highly developed and homologous with those of the rock-pigeon. " One point here should not escape attention, namely, that the transitional steps in Ocyphaps form a linear series; but there is nothing artificial or arbitrary about it. It is a small-number series, each element of which stands in an appointed place and marks the height to which the transformation process rose at that point in its course. Such a series can not be open to the objections which de Vries has very justly made against large-number series, the elements of which are collected at random and then arranged arbitrarily to display transitional continuity. In the Ocyphaps series there is some fluctuation, the series varying in length, but always advancing in one predetermined direction, like a tidal flow guided along a prepared channel, and flowing to varying distances, according to the initial momentum. 34 ORTHOGENETIC EVOLUTION IN PIGEONS. Here we find what at first glance looks like extraordinary variability (pis. 9 to 13) suggesting mutations, incipient stages, bars in statu nascendi, etc. The selectionist and the mutationist could each find what he looks for. The first thing to decide is the direction in which the phenomena are to be read. Is it a positive, progressive upbuilding of new characters, or a negative, retrogressive weakening of old characters? I have already anticipated the answer and will now briefly show how the direction of variation is decisively settled. (1) These spots have every outward appearance of being reduced remnants, such as we get in passing from the chequered to the barred condition in rock-pigeons. They are rounded or squarish in form, frequently irregular and thin at the edges, dull in color, as if fading, etc. (2) The smallest stages are not found on the exposed surface of the feathers, but lie concealed beneath the overlapping feathers next above or in front. Concealed spots admit Explanation of Plate 12. A. Wing-bars of adult stock-dove, Columba cenas, No. 5. x 0.9. A bird (young of 1904) captured by Mr. Lane. Bars of both wings. — Measurements of left wing. Hayashi del., Mar. 1905. Feather No. Total length. Exposed. Covered. Distance to tip. Remarks. First bar, 4 spots: mm. mm. mm. mm. 2 4 4 11 An extra spot on 2d 3 14 7 7 11 feather, amall and 4 19 10 9 14 obsolete; not seen 5 9.5 7.5 2 23 on right aide. Second bar, 5 spots: * 3 9.5 1 8.5 10 4 13.5 2.5 11 14 5 13 3.5 9.5 15 6 12 4 8 18 7 9 2 7 21 * On right only 4 spots, none on fourth, leaving an interval. Thud bur, (5 spots, all concealed; lower two thin and flccky ; found on third to eighth feathers; all rounded in form. Diameter of — First spot, 5 mm.; second spot, 8 mm.; third spot, 8 mm.; fourth spot, 6.5 mm.; fifth spot, 5.5 mm.; sixth spot, 3 mm. 7 spots on right side, 3 or 4 showing just at distal edge. No trace of a fourth bar in any specimen examined. (Later, in 1909, a specimen with a trace of a fourth bar was found; see pi. 13. — Ed.) B. Wing-bars of adult stock-dove, C. cenas, No. 11. x0.9. Hayashi del., Mar. 1905. A bird (young of 1901) captured by Mr. Lane. Bars of both xoings. — Measureme?its of left wing. Feather No. Total length. Exposed. Covered. Distance to tip. First bar (3 spots) : mm. mm. mm. mm. 3 9 2 7 14 4 19 7 12 15 5 14 8 6 20 Second bar (4 spots): 3 2 8 2 2 8 10 14 4 5 13 1 12 17 6 10 2 8 20 In several specimens first and second bars are nearly concealed as in this bird. Third bar (2 spots, rudimentary). — Concealed, thin, rounded, 4 to 5 mm, in diameter, long by 2 mm. wide, both very weak. Three spots on rigid wing. Second spot, 4 mm. PLATE 12 A. Wing-bars of adult stock-dove, Coluraba venas, No. 5. X 0.9. Hayashi del., Mar. 1905. A bird (young of 1904) captured by Mr. Lane. B. Wing-bars of adult stock-dove, C. cenas, No. 11. X 0.9. Hayashi del., Mar. 1905. A bird (young of 1904) captured by Mr. Lane. AMimSrnRjirirtiin PLATE 1 3 h * ■ • # • if \ %$ Rudimentary fourth wing-bar in adult female stock-dove, No. 24. Photographed Jan. 1909; hatched 1901 or 1902. The spots of the right wing wen- almost whollj concealed on the live bird; the second bar of the left wing was plainly displayed, its spots are plainly longer and stronger than in corresponding bar of right wing. A fourth bar (upper figures) is unmistakably present here. THE PROBLEM OF THE ORIGIN OF SPECIES. 35 of but one interpretation. This pigeon is a not-distant relative of the rock-pigeon, has a similar gray ground, and is therefore probably moving in a parallel direction, only more advanced. (3) The spots are found at the posterior end of the wing, near the upper edge, on one to three tertials and on a few long coverts. In some cases they occur also on a few of the second row of long coverts, 15 but here they are always very small and completely concealed. They are thus in the position occupied by vanishing spots generally. (4) The adult plumage makes no advance in the number of spots, and some spots (second row of long coverts), visible in the young, are completely concealed in the adult. This indicates degeneration unmistakably. (5) The stock-dove, although sometimes having a concealed third bar of few spots, never appears in chequered dress. It seems to have moved so far in the opposite direction that no reversal of course is now open to it- Taking the chequered pattern as the earlier one, the various conditions of chequers and bars in rock-pigeons, domestic races, and indeed in all the wild pigeons, become almost self-explanatory. We could not explain satisfactorily how just two bars could arise de novo in one species, three in another, twelve in another, and so on. The repetition of de novo origins would become ever more incredible. Making phylogeny our guide as to the starting-point, we find it comparatively easy to thread our way through the maze of patterns existing among 500 or more species of pigeons, and even to trace affinities farther back in the bird world. The orthogenetic process is the primary and fundamental one. In its course we find unlimited opportunities for the play of natural selection, escape the great difficulty of incipient stages, and readily understand why we find so many condi- tions arising and persisting without any direct help of selection. Charles Darwin: As natural selection acts solely by accumulating slight, successive, favorable variations, it can produce no great or sudden modification. (Origin of Species, Chapter XIV, p. 421.) Slight individual differences, however, suffice for the work, and are probably the sole differences which are effective in the production of new species. (Animals and Plants, Vol. II, Chapter XX, p. 233.) As modern geology has almost banished such views as the excavation of a great valley by a single diluvial wave, so will natural selection, if it be a true principle, banish the belief of the continued creation of new organic beings, or of any great and sudden modification in their structure. (Origin of Species, Chapter IV, p. 98.) August Weismann: The simultaneous modification of numerous co-functioning parts, in essentially different ways, yet in harmonious functional relations, points conclusively to the fact that something is still wanting to the selection of Darwin and Wallace. (Germinal Selection, p. 22.) We know of only one natural principle of explanation for adaptation, that of selection. (Ibid., p. 61.) The three principal stages of selection — that of personal selection, as held by Darwin and Wallace; that of historical selection as upheld by Wilhelm Roux in the form of a "Struggle of the Parts"; and, finally, that of germinal selection, the existence of which I have endeavored to establish — these are the factors that cooperate to maintain the forms of life constantly capable of life. (Ibid., p. 60.) 16 Rudimentary spots on the third row of coverts — a fourth bar — were later found (1909); these are shown in plate 13— Ed. 36 ORTHOGENETIC EVOLUTION IN PIGEONS. The harmony of the direction of variation with the requirements of the conditions of life is the riddle to be solved. The degree of the adaptation which a part possesses itself determines the direction of variation of that part. (Ibid., p. 54.) When a determinant has assumed a certain variation-direction it will follow it up of itself, and selection can do nothing more than secure it a free course by setting aside varia- tions in other directions by means of the elimination of those that exhibit them. (Evolu- tion Theory, Vol. II, p. 123.) Carl von Ndgeli: Between the theory of selection and that of direct causation, there is, apparently, only a little difference, since, according to the latter, the present condition of the organic world would likewise result from individual variation and elimination. But these two processes (selection and direct causation) differ fundamentally in their causal import. According to Darwin, variation is the germinating factor, selection the directing and regulating factor; according to my view, variation is at once both the germinating and the directing factor. According to Darwin, selection is indispensable; without it there could be no pro- gression, and organisms would remain in the same condition as at the beginning. In my opinion, competition simply removes what is less capable of existence, but it is wholly without influence in bringing to pass anything more perfect or better adapted. (Theorie der Abstammungslehre, p. 285.) The fortuitous or directionless variation of individuals would be conceivable, if it were conditioned by external influences (food, temperature, light, electricity, gravitation) ; for, as these causes obviously can not be brought into any definite relation to the more or less complex organization, they must effect sometimes a positive, sometimes a negative, step. If, however, the causes of variation are internal, in the constitution of the substance, then the matter stands otherwise. In this case the determinate organization of the substance must exercise a restricting influence upon its own variation ; and this influence, as develop- ment begins at the lowest point, can only take effect in an upward direction. (Abstam- mungslehre p. 12.) Individuals transmit to their offspring the tendency to be like themselves, but the offspring are not perfectly like the parents. The tendency to variation must therefore also be transmitted. A primordium, if all conditions are favorable, must be able to develop ever farther in a series of generations, as a capital enlarges to which interest is added annually; for each generation inherits from the preceding not only the possibility to realize the capital, but also the possibility to add the interest. (Individuality in Nature, 1856.) Hugo de Vries: According to the theory of mutation, species have not arisen through gradual selection continued for hundreds or thousands of years, but by steps, through sudden though small transmutations. In contrast with variations, which are changes advancing in a linear direction, the transformations to be called mutations diverge in new directions. They take place, then, so far as experience goes, without definite direction, i.e., in various direc- tions. (Die Mutationstheorie, Vol. I, p. 150.) CHAPTER III. THE PROBLEM OF THE ORIGIN OF SPECIES (CONTINUED). 1 In the Australian crested pigeon (Ocyphaps lophotes) we find the homologues of the two bars of the rock-pigeon, the elemental spots having nearly the same form, but a much more brilliant color (pi. 8). Moreover, in front of the secondaries and the long coverts which bear these two bars, each row of the median and lesser coverts of the wing bears a narrow band of black (text-fig. 11), and these, nine to ten in number, gradually diminish in width until at the anterior end they become mere lines. Between these bands and the two posterior bars there is the sharpest contrast in form as well as in color, the one being plain black and the other iridescent green and purple. Nevertheless, it can be shown that the two marks are serial homo- logues, two different forms, but one derived from the other. From the form and position of the bars, from the close resemblance they hold with the bars of the rock- pigeon, which resemblance is still closer in the juvenal feathers (pi. 8, fig. B, and text-fig. 10), and from the fact that bars among pigeons appear everywhere to be made up of like constituent elements, I feel certain of these homologues. The longitudinal spots composing the bars are, then, safely identified with the so-called chequers in the wild rock-pigeon and its domestic descendants, and both have been derived from the same old mark, named the dark center, of the turtle- dove pattern. There can be no doubt, therefore, as to which of the two marks found in the wing of the crested pigeon is the older element, and this relation furnishes the key to further orientation. The transformation of the longitudinal spots into the later transverse bands has already swept over a large portion of the field. Has the change been gradual and slow, or was it all accomplished suddenly, and in such a manner that no genetic continuity ever existed between the two marks? Were no inter- mediate steps or transitional phases required to pass from one extreme to the other? Was there no connection between them, and did one succeed the other as one picture replaces another by a turn of the kaleidoscope? Was there no predetermined direction of change manifest at any time before fulfilment? If orderless and directionless, then its appearance in the anterior rather than the posterior part of the wing, and its complete regularity in this part of the field, have no particular significance. These questions answer themselves in a most decisive way in the present case. We shall see that the change is a progressive one, still moving on in a definite course, which is premarked by gradual transitional phases. If these bands and bars were not found in one and the same bird, but one in one bird, the other in another, no one could hesitate to pronounce them distinct specific characters, and the mutationist would undoubtedly claim at least one saltation between them. But nature has here put these characters in such relations, joined them by such indubitable connecting stages, and now repeats her steps in such gradual ways 1 This was written in 1905 as a continuation of the preceding address (Chapter II). It is here published for the first time. — Ed. 37 38 ORTHOGENETIC EVOLTTTION IN PIGEONS. in successive plumages as to completely forestall the saltation hypothesis. Not only is the direction of the change hitherto discoverable, but its future course is predictable. In fact, a few allied species are found in which the change has been carried farther and in the direction plainly anticipated in the species we are con- sidering. One species {Lophophaps leucogaster) , also a crested pigeon and confined to Australia, shows this same variation carried through the whole row of long coverts (text-fig. 9), and already moving down the last row of feathers in the wing. Only 4 of the elongated metallic spots are left in this last row, while 5 to 7 are to be seen in Ocyphaps. The number falls to 3 in another species (Lophophaps plum if era). For questions involving the vital principles of the prevailing theories of organic evolution, the usual difficulty is to find tests that are really conclusive. As the main question turns on the nature of the relation between two specifically distinct conditions of a character, one of which is known or conjectured to be a variation or a mutation from the other, it is generally assumed, as a matter of course, that Text-figure 11. — Adult female crested pigeon, Ocyphaps lophotes. x 0.8. Hayashi del., L898. Spots and bars of left wing. Compare with Juvenal crested (text-figure 10) and with examples of Geopelia. we must go to two distinct but allied species for the two conditions to be com- pared. Variates are collected from both sides in order to see if they can be arranged in a single series running so continually from one species into the other that it will be impossible to find the dividing-line. Such series can be readily formed in number- less cases, but, as de Vries has so well shown, they supply no test. De Vries states : The overstepping of the limits occurs only in single and relatively rare individuals; the great majority belong to the mid-type of their species. If therefore we do not search for transitions, or if we do not seek merely to complete the series, but rather to make the measurements as numerous as possible, then the curves will come to light. Precisely that will come -out which we noted in the consideration of the Oenothera flowers. The well- known multimodal curves studied by Bateson, Ludwig, and many other investigators will appear. Every modal summit corresponds to a group of individuals belonging together— a type, or an elementary species. The transition-forms are easily detected by their infrequency. It becomes evident at once that they only apparently obliterate the limits, and that they can not possibly bring into confusion the renters of greatest frequency. They prove nothing beyond the fact that neighborin g curves on the same abscissa may overlap each other with their limbs. 2 2 Die Mutationstheorie, I, p. 307. THE PROBLEM OF THE ORIGIN OF SPECIES. 39 De Vries dwells at length and with effect on this " transgressive variability," as he terms it, endeavoring to make it abundantly clear that it is not really transi- tional, and hence can have no value as a support for natural selection. The trouble with such series is that the sequence is always wholly artificial and of doubtful significance, as it holds only for the extremes of single characters. Moreover, if the extremes are taken from two distinct species, they can not of course be interpreted as transitions looking to the origin of a new species. Such prospective significance can not lie between two species already in separate existence. It does not follow, however, that any sharp line of qualitative distinction is to be drawn between such variation and the variation (or mutation) which eventuates on new species. In statistical attire, individual variations may seem to be amenable to Quetelct's law, and to merely oscillate about a center which sets limits to their excursions, but never follows them. Thus conceived, they may appear to be quali- tatively different from variations that result in a shifting of the center of gravity to a new face of the polyhedron, to use the famous simile of Galton. If gathered at random and arranged arbitrarily, we may get as much chance out of them as we put into them. But is it safe to conclude that the mathematical theory of chance really covers all that is essential for a theory of such variations? It is on this assumed distinction that the whole theory of mutation rests. If true, then natural selection is not directly concerned in the origin of new species. Neither is there any room in such work for definite and continuous variation advancing in one direction. No matter how small the difference between two allied species, that difference is absolute and unmediated. Neither natural nor artificial selection has any power to raise the species to a new specific level. The species may be improved within its sphere, but the limit of such improvement is soon reached, and retrogression to the original status is swift and inevitable if the species is left to shift for itself. In the same way all human endeavor and progress is tethered, with no possibility of escape through cumulative improvement. Stare super antiquas vias is the utmost hope of the species. As de Vries remarks : A hard, a difficult, I should be inclined to say, a sad law. What is obtained with much trouble and care can be preserved only by as great care and trouble. And history is witness that this law is also true for mankind. Everywhere and always progress, but followed by regression as soon as the effort ceases. 3 Conclusions of so far-reaching import, based upon twenty years of observational and experimental research that has not been equaled since Darwin's time, certainly challenge our deepest interest. The distinction which threads this great work, binding it into one logical whole, is one not easily brought to a decisive test. The behavior of species in hybridiza- tion strongly suggests qualitative differences, such as are not apparent in ordinary individual variation. The usual aversion of one species for another, and the general infertility and instability of hybrids, seem to show that the behavior of species towards one another is influenced by something more deeply seated than habit. Ordinary individual variability fluctuates in such an aimless way, and is so unstable even when nursed continually in a definite direction and for long periods, 3 "Unity and Variability." Inaugural address, University of Amsterdam, Jan. 8, 1S0S. Also, The University Chronicle, Berkeley, Cal., Sept. 1898. 4 40 ORTHOGENETIC EyOLUTION IN PIGEONS. that it seems to stand in sharp contrast with changes of a specific nature, which, according to de Vries, are fixed on their first appearance. In biometric dress, individual variations present themselves as oscillations to and fro in a linear direction, while mutations appear to diverge in various directions. On the one side, every advance is sooner or later lost in regression; on the other, regression never appears. Variability seems to rise and fall with the food-supply, while mutation is not so affected. Mutation appears with the suddenness and completeness of a chemical substi- tution, and it can no more pass into a new specific form through flowing variation than could a crystal be slowly transformed into another of a totally different nature. In such respects, variability is represented as always standing at the opposite pole; cumulative to a degree, but without permanency; plus or minus, but never qualitative. The mutation theory puts the origin of species beyond the reach of investiga- tion, shrouding it in the utter darkness of premutation stages. The idea of fixity as an essential characteristic of species bars the way to accounting for the origin of the most remarkable phenomena of the organic world, namely, adaptations. The characterization of variability as linear is too superficial to be of diagnostic value. If fertility is a thing of all degrees, and that, too, within the limits of a single species, are not the qualitative distinctions inferred therefrom likewise things of degrees and variable within the same limits? If fertility may vary individually, as is certain, is there any reason to doubt that its physical bases are equally variable from individual to individual? But fertility varies also according to age and con- ditions of food, temperature, etc. 4 How then can qualitative differentiations be an infallible criterion of species? But discussion of such questions is somewhat aside the main purpose here. It is more profitable to take the fundamental issue to the test of facts. This issue is found in the denial of transitional stages in the evolution of species. If indubitable transitional stages between indubitable specific characters can be found, even in a single case, the mutation theory falls. To catch a character in the very act of giving birth to a new character, through gradual transitional steps, to be able to see the genetic continuity before it is broken for the last time, that is the kind of evidence most to be desired. Such evidence can only be given in a natural series, which is presented in a field completely oriented. The extremes of the series must be two characters specifically distinct, but not yet completely separated. The transitional stages must lie between these extremes, and lead in a definite direction and continuously from the later into the older character. Such transitions will gradually descend in the same direction; i.e., they will depart less and less from the earlier mark as they approach and pass into it. Under such conditions chance is not the law, and probability curves lose their enchantment. All the bearings are given by direct inspection, and no dispute can arise as to the transitional nature of the connecting stages. All these crucial advantages are placed in our hands in the species to be examined, and they are all the more decisive in that the transitional phases are to be seen in several consecutive stages of evolution in the young and adult birds. 1 The data which establish these facts are given in Vol. II. — Ed. THE PROBLEM OF THE ORIGIN OP SPECIES. 41 The orientation of the field is given in all essentials by simple inspection of the wing, and hence I need not here go into the genealogy and the present status of the older character, beyond the incidental mention of one or two facts for the sake of perspective. The elemental units of the two color-characters take, of course, the serial order of the feathers bearing them, and so fall in transverse rows. The sequence in each row is, again, that of the feathers, which run above downward. The sequence of the rows, as wholes, runs from before backward, and this antero-posterior sequence coincides, as will be seen, with the direction of the transformation of the color- pattern. As this direction is that of axial development in the embryo, it seems probable that it is determined by the same general laws in both cases. The pathway of transformation to be followed in the future is already pre- delineated, in part, by transitional phases, and hence the nature and position of the goal may be forecast. With the orientation indicated in a preliminary way, we may turn to the exami- nation of the several plumage stages. The simplest stage for this species (Ocyphaps lophotes) is to be seen in the first or juvenal feathers. It is here that earlier ances- tral features have not been entirely obliterated ; and for the sake of a better per- spective these may be briefly called to mind without citing evidences to be dealt with elsewhere. The ancestral pattern is that of the typical turtle-doves (Turtur turtur, pi. 2, Vol. II; wing, pi. 22, Vol. I) and T. orientalis (pi. 1, Vol. II, and text- fig. 3, this volume). It is everywhere among pigeons, as among many other birds, the point of departure in the evolution of color-patterns. The common element of this uniformly mottled pattern consists of a black spot occupying the center of the feather and a light border at the edge of the feather. The dark center has become pale in the young crested pigeon, and is indicated only by the dotted area in text-figure 10. The light edge is everywhere present, but it is also paler than the same part in the ancestral pattern. To this pattern two quite distinct new patterns have here been superadded. One of these consists of longitudinal black spots, or chequers, placed on the two posterior rows of feathers. The other consists of transverse stripes or bands, seen in all the anterior rows. In these longitudinal spots and transverse bands we have the two distinct characters to which we have looked forward. They are so different that at first they appear to be quite independent of each other. The two rows of chequer-like spots will readily be recognized as homologues of the two bars seen in rock-pigeons. Such spots are widely distributed among both wild and tame species, and although they have many specific forms they have everywhere the same origin from the dark central spots of the turtle-dove pattern. There can, then, be no doubt as to the specific value of these spots. That the bands are equally entitled to rank as specific marks will not, I believe, be questioned. Much smaller differences would pass as specific, if the marks stood alone, each in a species by itself. If one of two marks contrasting as sharply as these do can be transformed gradually into the other, it is of no vital importance how we rank them. The proof of transitional variation remains the same. It would be but a specious quibble to maintain that two such different characters can not be regarded as specifically distinct because they stand side by side in the same species. It can be seen readily 42 ORTHOGENETIC EVOLUTION IN PIGEONS. that one of these characters differentiates from the other; and how far would such differentiation have to advance before becoming specific? If carried to the extent of turning the two bars into bands, would the specific rank of the bands then be indisputable? If so, the whole point is conceded, for the transforming process advances to the end by transitional stages, as is plainly to be seen in the few species that have reached a higher level than Ocyphaps. Glancing now at the pattern as a whole, we notice first of all that the two charac- ters are separately distributed, each in a field of its own. As the rows of feathers succeed one another in regular order, without a break, how does it happen that just two rows have been reserved for the older character, while all the others are monopolized by the later character? There must be some definitely directed process of transformation to account for such relations. Is it a process that operates progressively and with continuity, or is every advance a clean-cut jump? But a linear series of jumps running in one direction would leave no very certain distinc- tion between mutation and orthogenetic variation. In the fore part of the wing the bands are very narrow and pale, and they completely vanish a little behind the front edge of the wing. As we pass backwards, they become wider from row to row, the last band being about twice as wide as the one just before it, and about one-third as wide as the bar that follows it. All this looks as if the transformation had moved along quite gradually, following the law of reduction of pigment, which I have demonstrated in other pigeons. If the nature and direction of the variation are what they seem to be, then the regional distribution of the two characters and the gradual increase of the width of the bands as they run backwards towards the bars begin to clear up. 5 However, before all the features and relations in the pattern and the nature of the variation can become fully intelligible, the mode of origin of the bands must be precisely ascertained. In this matter we come squarely to the main issue: Is it mutation or transitional variation? At first sight the juvenal pattern suggests mutation. Although the bands vary by increasing width, the difference between the last band and the bar following it is manifestly greater than is the difference between any two successive bands. But observe that in the upper part of the bar the bridge of transitions is given in three spots. It is here that we see just how the longitudinal elements of the bar are converted into the transverse elements of the bands. This is done by reducing the length of the spot, chiefly from the proximal end, until nothing remains but a narrow band-like remnant, with its longer axis at right angles to that of the original spot. 6 It would be very interesting to know whether the process is one of continuous development or one that moves mutation-wise, jumping all at once from the juvenal to the adolescent stage. As feathers once formed are fixed struc- tures, we can not of course expect to see one of these long spots reduced in length from day to day, and thus gradually turned into a transverse band. We can, however, demonstrate experimentally that the advance from the juvenal to the adolescent pattern is a progressive development that continues for at least several months and possibly for a year or more. Successive plumages and color-patterns seem like discontinuous phenomena, because they measure off at regular intervals the results of work that is done behind the scenes. The juvenal feathers show only how far the trans- forming process had advanced at the end of the fust week or ten days of life. The next plumage reveals the progress made from that time to the time of the first molt. The proof that the process moves on gradually from one plumage stage to the next is easily obtained by plucking a feather or two in the region of transitional change at any time or times before the first molt. In this way we may make room for new feathers to appear at any desired ages intermediate between the juvenal and adolescent periods. In these new feathers we shall get the juvenal pattern more or less modified in the direction of the adolescent according to the ages represented by them. I have not made this test in the crested pigeon, but have made it in the young of Geopelia cuneata with such perfect success (see Chapter X I that I feel safe in assuming that similar results would follow the test in other cases. PLATE 14 t £ ■ -/, $ .1 VI. \ 0\ £ vV ■• '"t/t .- t. , >x • "- Juvenal female crested pigeon, Ocwpfcajw lophotes, age 32 days. Toda del., Aug. 19( (posterior) three rows of feathers of left wing. PLATE 15 ' I - ^ Adult female crested pigeon, Ocyphaps lophotes. Toda del., July 1906. Last three rows of feathers of left wing. Compare with juvenal female and three stages of male. The "fundamental bars" are to be seen in all these feathers. THE PROBLEM OF THE ORIGIN OF SPECIES. 43 This simple mode of transformation is about three-fourths completed in the upper feather of the row, about half completed in the second feather, and only just begun in the third feather. If the figure (text-fig. 10) tells the truth, there can be no mistake about the conclusion. The direction and the manner of change are what we might have anticipated (compare also wing of adult female, text-figure 11). But is not this "bridge" a thing of chance that may appear once in a thousand times, more or less? And may we not find it next time in another place and telling a different story? Even the upper spot here is not quite reduced to a typical band, and may it not be a mistake to identify it as a stage of transitional significance? For certainty on this point we require a longer series, with the steps ending in a finished band. We should find the series invariably in the same place and running always in the same direction, in both sexes and at all ages. All these desiderata are amply fulfilled. The color-pattern does not reach full development in the first plumage and the male carries the development somewhat farther than the female. The female presents two and the male presents three consecutive stages in this development. Although the stages run nearly parallel in the two sexes, the female not only has one stage less than the male, but her two stages represent somewhat lower grades than the corresponding stages of the male. Hence we have the transitional series in five stages, which ascend in the following order: (1) juvenal female (pi. 14); (2) juvenal male (text-fig. 10); (3) adult female (pi. 15); (4) adolescent male (first- row coverts compared in pi. 17); (5) adult male (pi. 16). 6 Two different sequences are here to be distinguished. One of these falls within a single row of feathers in the individual plumage and is represented in the series of transitional steps leading from the new into the old character. Its closeness and length vary considerably from sex to sex and according to age, but only slightly in different individuals of the same sex and age. The second sequence runs parallel with the development of the bird, from the stage of the first plumage to that of full maturity. It is not a sequence of transitions, as presented in a single row of feathers of a single plumage, but a sequence of the stages of the same transitional series, as presented in successive plumages. The accompanying diagram (text- fig. 12) will illustrate the relations of the two sequences in the anterior bar in the three plumages of the male. The single feather with its transitional mark is the unit of the first sequence; the transitional series as a whole is the unit of the second sequence. The first sequence has continuity serially stamped upon it by nature; the second sequence 6 Before we go on with the examination of these stages we must take note of the new character which invades the field occupied by them. This new character is a brilliant metallic luster — a bronzy green in the long coverts, where we meet the transitional phases, and a metallic purple in the secondaries or last row of wing feathers. This iridescence, evidently one of the latest acquisitions of the species, is increasing, and spreads from the tip of the feather inward, in a direction precisely opposite to that of pigment reduction. It appears first of all in the field of the black spot, but spreads inward beyond the limits of the spot. The outline of the spot disappears in the full blaze of iridescence, but it may be readily brought into view by turning the feather until metallic reflections no longer reach the eye. This can be done most easily with feathers mounted in serial order on a card-board. Held at the proper angle, the outlines of the spots become clear, and we have before us a full series of transitional phases. The sequence of stages is likewise a continuous whole in which the stages are merely favored terms in a flowing series of intergrades. Iridescence thus appeals to be a phenomenon tending to elevate the spots and bring them within the sphere of utility. It seems not only to put a check upon the reduction of pigment, but also to actually turn the tide in the opposite direction, for the reduction in this region is not carried so far in the old as in the young male and the female, as we shall presently see. As the acquisition of metallic brilliancy is accompanied by an exceptional love of display in the male, the chief directing factor in its development may well be natural selection. 44 ORTHOGENETIC EVOLUTION IN PIGEONS. is broken into stages that represent results in the progress of differential develop- ment which are unfolded at successive epochs. The intervals between the stages give no outward indications of steps connecting them. Are we to conclude that this sequence is one of discontinuities? Although the transitional series linking two characters in one and the same plumage is the main theme in this paper, it will not be without interest to make clear that mutation has just as little to do with the second sequence as with the first. But first let us finish the examination of the stages individually, as this is necessary to complete the story of the transitions actually realized in each of them. The juvenal stage of the male, as we have seen, presents three transitional steps or intergrades between the two characters. I have not thus far examined the corresponding stage of the female, but it is safe to assume that the contrast between the last band and the first bar is less than in the male. The third stage in the order above named is that of the adult female (pi. 17, third row). In the uppermost of the long coverts we find a weak transverse band on the lower web. In the second feather we see just a shadow of a band on the upper web, and on the lower web a rather dim spot, in a half-way condition between the Jv.-• Ad.c? ra 2»- I i l I i V I I I i i I i y i v Text-figure 12. Diagram to illustrate pattern sequence in the anterior wing-bar of male Ocyphaps. 1, 2, 3 represent 3 feathers of the an- terior liar. I, II, III represent 3 plumage stages of the bar in the male — juvenal, adolescent, and adult. 1-3 show the first or feather sequence of transitions. I 1 1 1 show the second or plumage sequence of the transition s-ries. two characters. In the third feather the longitudinal spot comes to fuller develop- ment on the lower web. If the feather be turned just enough to cause the irides- cence to vanish, we see more plainly the shape of the spot and notice in its slanting proximal border a first step in the transformation to the band condition. In the next three feathers the spot elongates somewhat, but shows no sign of transforma- tion. The bronzy iridescence is strongest in these mid-feathers of the row and vanishes wholly on the tenth feather. In the fourth stage, seen in the adolescent male (pi. 17, first row), the transi- tional steps are extended through the whole row of spots, if we accept the last trace of a spot seen in the ninth feather. The series then consists of no less than eight steps in a single row of nine spots. No one can look at this series and fail to see that each step is a demonstration of transition, and that the series, taken as a whole, puts the genetic continuity of the two characters beyond a doubt. Here nature, as if to forestall the indirections of biometric gymnastics, draws a straight line between her two extremes, thus disclosing in a small number of steps a path- way of variation essentially different from anything within the purview of the large-number curve. PLATE 16 I I I I I ~'*.? Adult male crested pigeon, Ocyphaps lopholes. X 0.7. Toda del., May 1906. Last (posterior) four rows of feathers of right wing. The extent of iridescence is marked by dotted circles; the presence and depth of pigment by stippling of various degrees. THE PROBLEM OF THE ORIGIN OF SPECIES. 45 Let us follow this series through, step by step, beginning with the uppermost feather of the row. In this feather we find a well-defined half-element of a band crossing the lower web a little obliquely. On the upper web only a mere shadow of a counter half-element is discernible; and this is not discoverable in the juvenal stage and is only slightly strengthened in the adult stage. The half-element of the lower web deserves closer examination, as it represents a fully finished transformation. It has an even width of 1.5 mm. and sets off a light apical margin of the feather 5 mm. wide. In so far it comes fully up to the type. It falls a trifle short only in one particular, namely, its inclination from a transverse line is about 23°, which exceeds by at least 4° or 5° the slant of any cor- responding part in front of it. Slight as this deviation may appear, it is significant in the last step of a series that approaches the type gradually in this as well as in all other respects. This step is led up to, not only through the bar, at the upper limit of which it stands, but also through all the bands in front of it. To understand its peculiarity, we must therefore approach it from both directions — i.e., from the typical band as well as from the typical element of the bar. On the fore part of the wing, where the type is best realized, the two half-elements in each feather coalesce at the shaft-line and together form one straight mark crossing the feather at right angles. This rectilinear form prevails back to the last two or three bands, although a few elements may be found farther forward with a slight curvature; and in such cases the convex side of the element looks forward, and a minute notch is usually seen at the middle, making evident the double nature of the element. As we pass backwards, the width of the bands increases very gradually, rising from about 0.5 to 3 or 4 mm. In one out of a dozen cases examined a maximal width of 5 mm. was reached in the fifth feather of the last row. It is in this row that we get the widest elements, the strongest indications of their double nature, and at the same time an inclination in the half-elements that almost reaches the condition seen at the upper extremity of the bar. But these modifications in the band are so small that, viewed as a whole, there is nothing, at first sight, to indicate departure from the type. In the closed wing only the lower web of each feather is visible, the upper web being covered always by the overlapping lower web of the next feather above. The whole band, as thus presented, consists only of the lower half-elements, which flow into each other with such regularity as to mask the slight deviations from the type. For a clear view of all parts of the band the feathers must be plucked and separately mounted in their proper sequence (as in pis. 14 to 17). Thus exposed, it will be seen that the width, shape, inequalities, etc., although considerable, are yet so closely graded that the conditions in any one feather flow into those of the next above or below, so that the series may be read in either direction with equal facility. It is in the two upper feathers that the conditions approach more nearly those of the typical band, and it is here that we again meet with a slight inclination of the half-elements in the same sense, though not quite to the same degree as we noted in the half-element at the upper end of the bar. This inclination runs through the whole series and is most strongly expressed in the wider elements, which have a crescentic shape, with the horns following the outline of the feather and so curv- 46 ORTHOGENETIC EVOLUTION IN PIGEONS. ing forward. At the junction of the two halves of the crescent we have a broad re-entrant angle on the distal side, which vanishes as we run up to the series and sharpens into a minute notch as we go in the opposite direction. Throughout this row the lower half of the crescent is a little wider than the upper half, and this extra width (0.5 to 1 mm.) is added evenly to one side, namely, the proximal concave side. The result is that the lower half preserves its symmetry with the upper half on the distal side, while on the proximal side its curve is retained, but advanced by 0.5 to 1 mm. beyond that of the upper half, thus producing a sharp salient angle in the shaft-line, opposite the re-entrant angle of the distal side. The lower half differs also from the upper in having a slightly stronger inclination. This difference is quite small, but once recognized in this band it may readily be detected in lesser and vanishing degrees in the preceding band. Now, every one of the peculiarities here pointed out — the greater width of the posterior band and of the lower half-elements in this band, the crescentic form of the elements, the distal notch and the proximal angle, the inclination of the half- elements increasing with their width — all these peculiarities are so many transi- tional phenomena, passing phases of a transformation still in progress. The process, although always flowing onward in one direction, advances more rapidly in the upper than in the lower part of the field, and hence the crest of the wave, if such a metaphor be allowable, sweeps over the feathers of each row in regular sequence from above downward. This accounts for the differences in width between the upper and lower half-elements, for the approximations to the typical band being closer in the upper than in the lower parts of the rows, for the obliquity of the band elements, and for the gradual rectification of the latter as the transfor- mation comes nearer and nearer to completion. 7 Having seen that the first step in the anterior bar has peculiarities which are partially intelligible when approached through the bands that lie in front of it, it remains to examine its relations in the bar itself. In the second feather we find (pi. 17, first row), in the lower web, an oblique mark, strongly band-like in appearance, but in several respects falling short of the type to a slightly greater degree than the corresponding mark in the first feather. Its inclination is 29°, 6° greater than that of the first mark; its distance from the apex of the feather is 2 mm. greater; and its outline is not so sharply defined, the pigment spreading a little on either side. These differences, though small, are not to be credited to chance, for they are never reversed or wiped out in fluctuations from individual to individual. In the upper web is to be seen just a shadowy indication of the counter-mark, barely stronger than in the first feather. The proximal edge of this upper mark falls 2 mm. nearer the tip of the feather than the lower mark, and thus we get a lack of symmetry of the same nature but more 7 This change from an oblique to a transverse position, in which the rotation of the element follows the direction of the hands of a watch when we face the right wing, or the contrary direction when we face the left wing, is a phe nomenon that would repay further study. Is the rotation towards a position of more perfect equilibrium, depending upon the ontogenetic distribution of pigment in a symmetrical structure? Why, then, in the fore part of the wing, is the rotation carried beyond the vertical or transverse line, so that the half-elements lean slightly forward? And why in so many birds, with essentially the same symmetry of feather, are the half-elements permanently inclined so as to form V-shaped figures pointing sometimes backward, sometimes forward? These are only a few of the great multitude of forms which take their departure from bars and bands. Feather structure may have much to do with the primary phenomenon of transverse bands; bill certainly it can not be responsible for ornamental figures in which many feathers may cooperate, contributing very unlike parts to a single design, with a symmetry overriding that of the single feathers, and following lines that regard only the form of the bird as a whole. PLATE 17 / / - The longer left wing-coverts of adolescent (second plumage) male, adult male, and adult female crested pigeons. X 0.9. Toda del., Apr. 1906. For comparison, feather by feather, in the three forms. Depth of pigment and extent of iridescence indicated as in pi. 16. The blue color gives the angles formed by the proximal border of the band-elements and the spots with the feather shaft (see text). THE PROBLEM OF THE ORIGIN OF SPECIES. 47 strongly expressed than that before described in the crescentic marks of the pos- terior band. On the distal side of the lower mark, or half-element, there is a very thin sprinkling of dark pigment, the meaning of which becomes clear in the next feather below. There is also a very weak iridescence overlying this mark and extending over the surface to near the apex of the feather. In the third feather the band-character falls still farther below the type, so far, in fact, that had we only this one feather to look at its identification would be some- what doubtful. Close comparison with the first and second feathers, however, is all that is required to make the conditions here intelligible and to let us into their full transitional significance. It is in this feather that the two characters — the bar and the band — are presented in nearly even balance, the former being near the mid-stage of decline, the latter at about the same stage of its rise. What appeals most to our interest here is the union of the two characters within the field of the older one, the individuality of the latter being remolded directly into that of the new character. We can not, of course, see the remolding, but only the several steps in their natural order. Let us look at the picture in this feather, then, as a delineation of a single step in the series — as an image of a character caught by the feather at a moment when the transmutation was in actual progress. We noticed that the inclination and width of the band-character and the distance from the tip increased in passing from the first to the second feather, while its sharpness of outline diminished, especially on the distal side, where a sprinkling of pigment appeared, spreading away to near the edge of the feather. These peculiarities are increased or diminished in the same respective senses in pass- ing to the third feather. The sprinkling of pigment is strengthened to a degree that enables us to recognize in the picture as a whole the form of the longitudinal spot seen in the corresponding juvenal feather. In short, we see here our two characters blended into one; or, more accurately speaking, the new band-character arises from the proximal border of the original spot, or bar-character, and hence takes its inclina- tion, outline, and distance from the tip of the feather directly from the spot itself. The shape of the spot as a whole is given in plate 17 (first row). It is approxi- mately fusiform and is set obliquely just within the white crescentic edge of the feather. Its long axis is 12 mm. and its short axis 6 mm. Its distal side is strongly rounded; the proximal side is but slightly curved and inclines 36°. In the upper web a small triangular area of pigment is seen, its longer side (4.5 mm.) resting on the shaft. This is a remnant (not a rudiment) of a spot; the proximal border of which shows an obscure band-like differentiation (compare adult condition). The proximal border of the spot in the lower web represents the band-character rising into plain view in its lower third, but attaining only a shadowy outline in the upper part, where the pigment is sprinkled in only a little more closely than elsewhere in the spot. This border has a slight curvature, which becomes a little stronger towards the lateral edge of the feather, finally makes an abrupt bend, and runs out into a small and pointed extension, thus giving a rather sharp boundary to this angle of the spot. The extension reaches only 1.5 mm. beyond the border proper, and is therefore quite inconspicuous. It is constant, however, and becomes a convenient landmark for locating the band-character in the first and more obscure stages of its appearance. It sinks almost out of sight in the second feather, measur- ing here only 0.5 mm., and no trace of it is left in the first feather. 48 ORTHOGENETIC EVOLUTION IN PIGEONS. In the fourth feather we find at first glance no striking change in the general form and extent of the two spots. It is only by close inspection of the minute features that we discover that the trend of cumulative changes is still in the same direction, and that as the band fails more and more of its highest development, its individuality becomes less and less distinct from that of the bar. Both spots are a little more strongly and evenly pigmented than in the third feather. The proximal borders are less band-like, but fall only in fine degrees below the condi- tions seen in the third feather. In the upper spot there is still a suggestion of the band in the slight preponder- ance of pigment in the proximal part of the triangle. There is also a very slight incurving of the upper side of the spot, corresponding to the deeper curve seen at the same place in the third feather. In the lower spot the condition of the proximal border is particularly instructive. Although the spot as a whole is a little more strongly pigmented than in the third feather, the border takes a step in just the opposite direction; that is, it sinks plainly below its height in the third feather. The two characters are no longer in equipoise; the scales turn in favor of the spot. Let us compare closely, in order to see with what nicety the transitional degrees are marked. As shown in the figure, the loss of pigment is so nearly even for the entire proximal border that it still looks as if it were the same thing before described, except as weakened in quantity of pigment. As before, the pigment is densest at the lower end, becoming gradually thinner towards the shaft, until in the upper third its individuality, if it have any here, is quite beyond recognition. The con- ditions in the two spots have changed so very little that we have no difficulty in seeing just where the pigment should lie if present; but I find that no border is here marked off; band and bar are one in this third. The bend of the border around the lower angle of the spot and its pointed extension are still quite distinct, though not so sharp in outline as before. The distance of the bend from the tip of the feather is here 1.7 mm. greater than in the third feather. In the feathers thus far examined, the inclination of the border and the distance of the bend from the tip of the feather run as shown in the accompanying table. From these measurements it will be seen that the distance from the tip is closely correlated with the inclination of the proximal side of the spot, every 1 mm. in one column corresponding to about 2° in the other. As we descend the series, the band gradually loses its transverse position and sinks out of sight in the proximal border of the enlarging spot. As we ascend, relations are reversed; the spot is reduced in length, its proximal border becomes less and less inclined and more and more band-like. In one direction the band merges in, in the other it emerges from, the bar. Evolution is here linear, and hence it can have but one direction. » The figures for inclination given here do not agree, except in the sense of an orderly increase, with those given in connection with plate 17. — Ed. Feather. Inclination. 9 Distance. m m . First 11 6 18 10 Third 27 13.5 Fourth 30 15.2 CHAPTER IV. THE ORIGIN AND RELATIONSHIPS OF THE ROCK-PIGEONS AS REVEALED IN THEIR COLOR-PATTERN. 1 The wild rock-pigeons, universally regarded as the ancestral stock of all our domestic pigeons, exhibit two very distinct color-patterns, one consisting of black chequers uniformly distributed to the feathers of the wing and the back, the other consisting of two black wing-bars on a slate-gray ground. The latter was regarded by Darwin as the typical wing-pattern for Columba livia; the former was supposed to be a variation arising therefrom, of frequent occurrence but of no importance. Just the contrary is true; the chequered pigeon represents the more ancient type, from which the two-barred type has been derived. From this standpoint the chequered type deserves the distinction of the specific name, Columba affinis, once accorded to it but subsequently abandoned by Mr. Blyth. Columba livia could well be reserved for the derived type. The direction of evolution in pattern in the rock-pigeons has been from a con- dition of relative uniformity to one of regional differentiation. That seems reason- able enough, and yet Darwin's view to the contrary rested on plausible grounds. Indeed, the chequered wing might, at first sight, appear to be a departure from the more to the less uniform condition; and this would be true with respect to the gray surface, considered as a plain unadorned area, that originally included the entire wing. The appearance of two bars would be the first break in uniformity, and the extension of chequers, like those composing the bars, to all the coverts, might seem like turning uniformity into variformity. On the other hand, if we attend to the black chequers rather than to the gray ground, it is at once apparent that the chequered wing presents a condition of fairly uniform distribution of like spots, while the two-barred wing represents a long departure in the direction of regional differentiation. It is quite natural to suppose that an unspotted stage preceded the spotted one, but general theory of this kind is often misleading. Uniformity in color is not always proof of simplicity. Some of the more highly developed colors, as green and blue, may be quite as uniform as the gray of the pigeon's wing. Even iridescence may sometimes extend more or less evenly to large surfaces and at times characterize the plumage as a whole. The gray of the pigeon is certainly not to be regarded as a primitive ancestral color. Its evolution has not been worked out, but it is known that it is not a pure pigmental color. It is an effect produced by the withdrawal of brown pigment from the superficial parts of the feather, and by a segmental arrangement of the pigment granules in the barbules (Hacker). Gray in the pigeon is undoubtedly a relatively late acquisition, brown repre- senting the simpler and more primitive color. Evidence of this is seen in the 1 An address read November 18, 1903, to the National Academy of Sciences at the University of Chicago. The slight amount of repetition of statements made in Chapter II will probably not prove a disadvantage to the reader. Two pages of a manuscript (Z 10), written in 1907, have been added to amplify the briefer statements of the earlier address. A detailed comparison of the individual spots or chequers in Columba livia and Columba affinis has been included. The address of 1903 was also delivered at Woods Hole, and a one-page abstract of it published in the Bio- logical Bulletin, Volume VI (1904), p. 307. That abstract is used as a partial summary of this chapter. — Ed. 49 50 ORTHOGENETIC EVOLUTION IN PIGEONS. prevalence of brown among the wild species of pigeons, and, further, in the fact that in all species with a gray plumage in the adult state the juvenal feathers are always suffused with brown. The first ancestors of the pigeons, or of the pigeons together with all other birds, may have been of a single unspotted 2 brown color. But as color-patterns are found quite generally distributed among the lower orders of animals, it is hardly probable that the avian branch came into existence entirely unprepared for the decorative supremacy it was destined to enjoy. Be this as it may, it can be made very clear that the rock-pigeons are not those first ancestors, nor yet their nearest relatives among the pigeons. The rock-pigeons represent a terminal twig in the avian genealogical tree. The main stem of the pigeon branch, if I read correctly the testimony of color-patterns, is represented most nearly by the turtle-doves (Turtur orientalis and T. turtur) of the Old World. The original turtle-dove pattern, in which all or most of the feathers were similarly differentiated into a dark center and a light edge, seems to have been a veiy general if not a universal avian pattern. This pattern certainly preceded the chequered type of the rock-pigeon (Colwnba affinis), and it is possible still to find connecting types — types in which the turtle pattern coexists with the chequered pattern— the latter coming in to replace or supersede the former. Examples are to be seen in the bronze-winged pigeon (Phaps chalcoptera) of Australia, the Florida ground- dove (Chamcepelia passerina) , and some others. The genetic connections between these two patterns belonging to two distinct families have been considered in Chapter II. The main problem to be dealt with in this paper concerns the closer relation of two patterns differing at most only in specific rank. That the two patterns are connected by direct lineal derivation is not questioned. But what is the direction and what the mode of derivation? The problem is only one of thousands of similar problems; but if it admits of a decisive solution it will furnish answers to a few questions of leading interest at the present time. The wing-bars of the rock-pigeon represent a specific character with a history that can be clearly read in both the wild and the domesticated forms. The chief transitional phases are to be seen in almost any flock of mixed breeds. The direc- tion of evolution can be determined in two ways: (1) by comparison of individuals, young and old, in each of the two species and in domestic races; and (2) by com- parison of the different wild species that are marked with spots and bars in various stages of evolution or decline. The conclusions arrived at through comparative observation can then be tested by experiments in breeding, cross-breeding, etc. In tracing the origin and genesis of a single character we meet the leading questions in the evolution of species. First and foremost, the question as to the nature of the initial stages. Did the character arise as a variation de novo, or as a progressive modification of a preexisting character? If de novo, did it spring suddenly forth, with some decisive advantage in the struggle for existence? or did it appear as one of many minute changes, and by some happy chance get a start that gave it the lead in future development? In other words, did it begin as a discontinuous variation, sport, or mutation? or did it arise cumulatively, as a con- tinuous development? If it originated by modification of an earlier character, was 2 Later Professor Whitman questioned this possibility much more seriously than is done here. — Ed. THE ORIGIN AND RELATIONSHIP OF THE ROCK-PIGEONS. 51 it at first a sudden, sport-like departure? or was it a slow and continuous transfor- mation of a progressive or retrogressive nature? Then we come inevitably to the deeper question, which natural selection only partially penetrates — the question how variation, multifarious and undirected, without the aid of design or a designer, can advance to such definite and wonderful achievements as specific characters. As the bars and chequers represent the two known extremes of a lineal series of variations, if we could determine which came first in time, the problem would be fairly within grasp, for then the direction of modification would be clear, and the mode of progress or retrogression would be but a matter for simple observation to settle. Such series can usually be read in either direction with about equal facility, and even if all the intermediate terms are before us, the direction of sequence is still ambiguous. This ambiguity infects many a biological problem, and in the field of color-patterns it is omnipresent. In the present case it can be eliminated, or reduced to negligible limits, by taking bearings from a number of different vantage-grounds. Which is more primitive in the rock-pigeons, the chequered or the uniform color? The following detailed description of the chequers and bars of a bird representing each of these two forms assist in supplying the answer. 3 One of three chequered individuals received in 1898 from Mr. Chatwin, of Dover, was less thickly chequered than the two others, which were thickly so marked. I found in the least-chequered bird, which is here described, only eleven feathers, so that at least four or five are missing; probably in molt. The measurements are omitted, as the series is not complete. 4 One or two points, however, are worthy of notice: (1) The spots here are longer than in the uniform gray individual (pi. 2), and pointed behind. (2) The gray tips are here as much shorter than in plate 2 as the spots are longer — the spots being about twice as long and the tips about one-half as long. The spots being at near the mid-length of the feather in both sets, the gain in length in this set, or the loss in length in the other, is wholly an elongation towards, or retreat from, the tip. Assuming, as I believe one must, that the elongated pointed spot is the more primitive, we have the means of judging whether the uniform gray with two bars of more or less squarely cut spots, or the chequered variety with its second bar always more or less serrate, stands nearer the original type. I do not hesitate to place the chequered variety as the earlier, for its spots are of the earlier pointed form which is so generally characteristic of the chequers on the anterior coverts. The uniform gray with two bars is a derived, later pattern, and not, as supposed by Darwin and others, the ancestral form, from which the chequered type is a secondarily derived form. This view is fully borne out by all the facts we can obtain : (1) The most highly developed bar is the first (posterior). It is in this bar that we find the greatest width and the greatest depth of color. Passing from behind forward, we find the bar-spots less and less developed. This is true in all species with bars; e.g., in crested pigeons the posterior bar is the most highly colored, the second bar is less brilliant, while the third is still less so, and those anterior to the third grow narrower and duller. (2) The posterior bar seldom shows traces of a serrate edge; the second bar shows all stages of serration, from the long, sharp point to the obtuse, rounded, slightly convex, and 3 The following pages (MSS. A . r >0), in smaller type, have been inserted at this point by the editor. * An illustration of the chequers and bars of this bird was prepared, but it is not available to the editor. 52 ORTHOGENETIC EVOLUTION IN PIGEONS. straight-cut outline. The less black in the wing as a whole, the more serrate this bar, as a rule; the more black, the more the long point fills up towards the straight-cut form. (3) In the first plumage of young domestic doves I find that the second bar generally exhibits pointed spots, and more strongly pointed as we descend from the back — where the spots are most developed and most nearly straight — to the lower edge of the wing, where the spots disappear, or become much less developed and increasingly pointed. In the second plumage these spots become fuller and less pointed, often losing all marked serration. (4) The first plumage of passenger-pigeons and the adult plumage of passenger females contains well-marked serrate bars, which are more or less lost in the adult males — the males having a more nearly uniform color than the females and young. (5) The young of Geopelia humeralis have the long-pointed spots in the first plumage. In the second plumage they are wholly absent. (6) The fact that the element of the bars is so generally distributed among wild species and that it generally takes the pointed form, or a form plainly derived from the pointed type, is a strong evidence that the pointed spot is the archaic form. (7) On any other assumption, what explanation do we have for the pointed pattern in so many wild species, and especially in the first plumage of pigeons? Evidently we must have a view consistent with some sort of evolution of the bars. If the pointed spot is not the original element, then what is more so? (8) Two bars are nowhere indicated to be the original number; on the contrary, the evidences are for more or less even distribution of the spots over the wing and scapulars. It is, in fact, easy to find many different species with many bars in progress of evolution, but no species where these marks are limited to two, as in some of the Columba liria. In the other form — a barred wild rock with traces of a third bar — the tertials and secondaries (14 in number) graduate into one another in form and color-marks so that no dividing-line is possible (pi. 2). This (first) bar is compounded of two distinct bars, one being terminal (secondaries), the other subterminal (tertials). The upper, subterminal bar continues from the fourteenth feather to the fifth; the terminal from tenth to the first. The two, therefore, are present together on the tenth to the fifth — i.e., on 6 middle feathers — the lower 4 secondaries having only the terminal, while the upper four tertials have only the subterminal bar. On the middle six feathers (10 to 5), the terminal bar diminishes upward, while the subterminal diminishes downward. In some of my domestic pigeons both bars extend farther, the terminal running up on the tips of the tertials, the subterminal running down across the lower secondaries and even across the primaries in diffuse freckles. The extension in opposite direction is the later development. We have, in this (first) bar, double spots (upper and lower) on the upper four feathers and a trace on the tenth; the spots of the subterminal bar average about twice the length of the spots of the second bar. The subterminal bar alone is the serial homologue of the anterior bars, while the terminal bar is not represented in the anterior feathers except for a bare trace, as seen in young domestics. This trace is very interesting, showing that there is a tendency to repeat the same pattern in each feather, and if the pattern is reduced it is the weaker position that suffers first; e.g., the anterior bar disappears before the second or first; the lower parts fail of development sooner than the upper parts. Reduction of development leaves the pattern in a more ancient form — that is, in elongated spots, rather than square or rounded spots. In looking at the first bar of a young domestic pigeon, I notice that in the folded wing only the upper ten feathers — secondaries and tertials — take part in forming the visible portion of the bar (i.e., the subterminal bar). The lower four secondaries have the spots diminishing in extent and depth of color downward and each showing the original pointed THE ORIGIN AND RELATIONSHIP OF THE ROCK-PIGEONS. 53 form; but in the closed wing these spots are entirely out of sight, and only the terminal spots of these feathers and portions of the "gray spots" are in sight, and these are so closely folded as to fall nearly in line with the subterniinal bar, and thus appear at first sight to form a part of it. The part they appear to form, however, is very small, so closely are the feathers folded up under each other. The whole length (from top to bottom) of the first bar in this specimen is 65 mm. and the part formed by the lower four secondaries is not over 8 mm. — i.e., not over one-eighth of the whole bar. Now, in the above-described Columba livia (pi. 2) these four secondaries have no trace of the subterniinal bar, and the next two feathers have only freckled indications of the bar. In a wild female passenger-pigeon the "terminal spots" of some six or more secondaries form the lower half of the bar, and the upper half of the bar (subterniinal spots) is not longer or more conspicuous than the lower half. The second bar shows five spots belonging to the middle long coverts, the middle spots being the larger and longer, the upper and lower spots diminishing until only the tip of the spot is visible. The third bar consists of six or seven visible spots, and this bar comes mainly in the lower half of wing. The fourth bar shows two to four spots in the lower middle part of the wing. In front of the fourth bar I see from one to two or three spots at different points — elements of so many imperfect bars, extending up to the front end of the wing. This (female passenger) wing is a very instructive type, especially in a comparison with chequered homers. The distribution of bars and chequers among wild species of pigeons may be trusted to give unequivocal indications as to the origin of bars. There are several kinds of bars, all compound, and all composed of similar elements. They are found in only a few of the more highly developed pigeons, such as the crested pigeons of Australia, the red-breasted pigeon of the Philippines (Phlogamas luzonica), the stock-dove of Europe, and the rock-pigeon. The chequers composing the bars are everywhere the same, homologues of the chequers of the rest of the wing. These spots are very widely distributed, and in most cases there are no bars present to which their origin could be referred. The conditions presented by the bars are everywhere indicative, not of a simple, primitive character, but of a character derived from simpler elements by various modifications in form. There is no evidence that bars are the incipient stages in the evolution of chequers. They appear rather to have already reached their highest stage of differentiation, and in some cases to be in decadence. This is most decidedly indicated in the much-reduced bars seen in the stock-dove (see pis. 9 to 13). Conditions have been noted even in the rock-pigeons looking towards disappearance of the bars. 5 "Occasionally," says Darwin (Animals and Plants, Vol. I, p. 225), "birds are seen in Faroe and the Hebrides with the black bars replaced by two or three black spots; this form has been named by Brehm C. amalia?." The bars have been completely obliterated in some races of domestic pigeons and more or less reduced in width in many others. In comparing the juvenal with the adult plumage we get another class of evidences that are scarcely less than decisive in their bearing on this question. In some species chequers are present in the juvenal feathers and wholly absent in the adult plumage. This is the case in the small ground-doves 6 (Geopelia) of 6 In a few individuals of our band- tailed pigeons (text-fig. 5, Vol. II) I have seen minute blotches of black cm smiu feathers in the place of the bars; the affinities of this pigeon, as well as its high general finish, compel us to regard these flecks as the last vestiges or atavistic reminiscences of lost bars or chequers. 6 The geopelias are illustrated and further described in the following chapter. — Ki>. 54 ORTHOGENETIC EVOLUTION IN PIGEONS. Australia. On the secondaries and the longer coverts these chequers fall into regu- lar rows, conforming necessarily to the serial arrangement of the feathers hearing them. They are thus in a position and order corresponding precisely to the position and order of the elements of the bars in Colurnba livia. To convert them into typical bars it would be necessary only to enlarge them transversely until they coalesced by juxtaposition. These transient chequers, distributed to all the coverts of the wing, seem to tell us that the geopelias were descended from chequered ancestors, whose; color-marks are now recapitulated, but soon superseded by the Inter pattern which characterizes the species in the adult stage. The barred con- dition of the adult plumage differs widely from that of C. livia; but the important point, is that longitudinal chequers are first and transverse bars second in order of development. While all species of Geopelia exhibit the same transitory chequers in the juvenal feathers, in the adult plumage two distinct patterns are presented. In G. cuneala 1 the ground-color is gray and each of the upper wing-coverts is marked with two small lateral, roundish, white spots, which fall into regular transverse rows on the longer coverts. These rows cross the wing in lines that would be followed by transverse bars, and development shows that the spots are differentiated remnants of bars. In the other geopelias, narrow transverse bars are formed by the coalescence of the dark crescents at the tips of the feathers. As these bars follow the tips of the feathers, they are really composed of serially connected scallops that simulate bars only in a few of the longer rows of coverts, where the feathers overlap in a way to conceal the upper third of each crescent and leave the lower third exposed as com- ponent parts of nearly straight and continuous bars. In the ontogeny of the geopelias we see the different species all taking depart- ure from a common color-pattern, consisting of longitudinal spots or chequers distributed uniformly to all the feathers of the wing. We see this common and earlier pattern retained for only a few weeks, then exchanged for the specific patterns of the adult birds, the later patterns developing in two widely diverging directions, leading to two types of generic rank. In the geopelias the chequers are completely lost with the juvenal feathers; in a distinct but allied genus (Ocyphaps), the crested pigeon, an interesting variation is presented. 8 Here typical chequers have been retained in two rows of feathers — the long coverts and the secondaries — and replaced in all the other upper wing-coverts and scapulars with narrow transverse bars, differing from those of Geopelia in being subterminal and straight and in being directly derived from the original chequers by simple reduction of the latter in length. In the second row of coverts next in front of the long coverts — the reduction is less than in the anterior coverts, and consequently the bars are here more evidently shortened chequers. In the young bird the bars shorten gradually from behind forward, from chequers of full length to the narrowest anterior bars. The mode of trans- formation is demonstrated to perfection in the two inner feathers of the row of long coverts. 7 Reichenbach gave this unique species the generic name Stictopelia, and Bonaparte concurred in this. I think the development of the color-pattern warrants the separation fr the other geopelias. "This is fully described in the preceding chapter. -Ed. THE ORIGIN AND RELATIONSHIP OF THE ROCK-PIGEONS. 55 The two posterior rows of chequers, which have been preserved in this genus and raised to a state of brilliant iridescence, are the exact homologues of the two black bars of Columba livia. The starting-point is then again found in general chequers, and evolution has advanced by a definitely directed reduction of pigment, proceeding from before backward, 9 in the same direction as that of embryonic development. That this is the direction of evolution, and that there is no possi- bility of reading the conditions in the opposite direction, becomes still clearer when we discover that a closely allied genus (Lophophaps) has taken a step in advance of Ocyphaps, in that it has acquired the narrow bars in the place of chequers in the row of long coverts, leaving but a single row of chequers on the secondaries. That Lophophaps occupies the higher plane is obvious in other modifications seen in the head, breast, and wing-pattern. If the testimony of juvenal patterns be extended to species having chequers so closely agreeing in form, color, and distribution with those of the chequered rock-pigeon that no shadow of doubt can remain as to their homology, we shall get only additional confirmatory evidence that the process of evolution in color- patterns has been a sweeping one, involving the whole surface and taking the same general direction. The stages reached are various, ranging all the way from the full-chequered to the wholly unchequered state; from chequers and bars combined in different proportions to bars alone; from many bars to three, two, one, a remnant, or none; and in all shades of brown, black, gray, red, to pure white. Nowhere in this field of variations do we find any indications that chequers originated in the form of bars at the posterior end of the wing and then spread from behind forward. The wild passenger-pigeon (Ectopistes) bears chequers closely resembling those of the chequered rock-pigeon. 10 There can be no mistake here as to the direction in which the phenomena are to be read. The direction is as certain as that the adult male stands in advance of the adult female, and still more in advance of the young bird. The significance of the case lies mainly in the fact that it is not an isolated or exceptional one. Many other species tell more or less perfectly the same story. A parallel case, only carried still farther in the same direction, is found in the mourning-dove (Zenaidura). The adult male (see pi. 19, Vol. II) and female differ but slightly, each having about a dozen chequers visible on each side. These are confined to the scapulars and to a few feathers at the posterior upper edge of the wing. In the young (pi. 7) they are more numerous, but less so than in the young passenger-pigeon. The middle and fore parts of the wing in the adult have no visible chequers, but a few concealed ones which may be seen on lifting the over- lying feathers. These concealed chequers and other differences between old and young show that the species had its origin in a chequered stock and that its history has been analogous to that of the passenger-pigeon. The reduction has not been carried out in a way to leave bars, but the location and the clean-cut outlines of the spots, as well as their intensity of color, suggest that some influence may have operated to preserve and then intensify them in a region suited to their most effective display. 9 Just the reverse of the contention of Eimer and his school, that evolution in color-marks progresses in a postero- anterior direction. 10 These have been described and illustrated in Chapter II. — Ed. 5 56 ORTHOOENETIC EVOLUTION IN PIGEONS. It is here that we may with some reason suspect the intervention of natural selection. It would, in this case, come in, not as a primary factor to originate a new character, but adventitiously, by invitation, as it were, of favoring predetermi- nations and environmental conditions. The process of deletion of the chequers, traveling backward and upward, would tend directly to clearing the field in a way to leave a few chequers just where we now find them. The ornamental value of these few chequers and their utility as recognition-marks would obviously be enhanced by their isolation in a plain ground, just as a few trees, concealed in a large forest, become conspicuous when left standing alone. These chequers, being on the larger feathers, would have the advantage of size, and so their preeminence, ■ w^fe- a^. A J Text-figure 13. — Adult male Nesopelia galapagoensis. Natural size. Hayashi del., June 1907, from a skin loaned by the U. S. National Museum. 1. Three or four irregular rows of the anterior lesser coverts have lost the chequers. 2. Following these are one or two rows of lesser coverts with vestigial chequers, mostly concealed. 3. Then follow five rows of coverts bearing chequers. Of these five rows, one could be reckoned with the lesser coverts, three as median, and one as long or larger. Each feather usually bears two spots, the outer spot the larger; passing from the inside outward, the spot on the inner web becomes rapidly smaller until at least it is only a marginal streak, and then the outermost feathers bear only one chequer on the outer web. The chequers are mostly rounded ciistally. The dotted part of the feathers is brown, and the white spaces between spots and behind them are left as they appear. 4. The scapulars are double-spotted. 5. The tertials are double-spotted, but the outer spot becomes smaller passing downward. 6. On the inner tertials there is asymmetry in position of the two spots, as in mourning-dove. 7. The secondaries have black spots, so that here we get the essentials of a posterior wing-band. 8. Many feathers of the back, between the scapulars, are spotted black. Nesopelia is more primitive than the mourning-dove in the number of spots, but has white between spots, which makes them much more conspicuous than in the mourning-dove. attained without the aid of natural selection, would be an open door through which it might enter and contribute to their improvement. The part possibly taken, however, could at most be but a late and inconsiderable share of the total achievement summed up in these spots; and the course of events in at least one of the allied forms (Melopelia) indicates that these marks are destined to be washed out. The question here raised is one of general interest, and it could probably be settled by an extended study of the marks in the mourning-dove and a comparison of all the Zenaidinae. Nesopelia galapagoensis is the only one of this family which is fairly evenly chequered over the whole wing and scapulars (text-fig. 13), and this condition stamps it as the oldest member. Melopelia, with its two species, THE ORIGIN AND RELATIONSHIP OF THE ROCK-PIGEONS. 57 stands at the opposite extreme, with not a single black chequer left on the wings, and is therefore to be classed as the latest branch of the family. Intermediate stages, with few to fewer spots, are seen in the two genera, Zenaida (pis. 83 to 87) and Zenaidura (pi. 19, Vol. II). The white-winged pigeon {Melopelia leucoptera) is a most instructive form. Although a much more highly accomplished bird in the arts of display of form, feathers, and voice than the mourning-dove, it has suffered a complete effacement of the chequers it once possessed in common with other members of the family. Indubitable proof of this is to be seen in the juvenal feathers, which, in some cases, exhibit a few pale vestigial chequers (text-fig. 14) in the last row of long coverts at points where the chequers are usually best developed in chequered species. Text-figure 14. — Left wing of juvenal white-winged pigeon, Melopelia leucoptera, age 25 days. Natural size. Hayashi del., May 1899. Rudimentary (obsolescent) spots are shown, but so nearly faded out and so nearly covered that one would not notice them unless looking especially for them. Color dark gray; in shape they are pointed posteriorly. The spot is elongated along the lower margin of the feather. There are only a few of them; three or four of the longer tertials have long, narrow edge streaks, completely out of sight under the longer coverts. I found four spots on the upper longer coverts, the lower of which was covered and is not seen in figure. The next row of coverts has three or four spots, rather smaller and more obscure, at about the same level as spots in long coverts. The secondaries are quite dark, with white edge and tip. This is the lower half of the first bar. The upper half of this bar is represented by the three or four covered spots of the tertials. I could find no other spots on the coverts or scapulars. When the feathers are mounted on white board, one can scarcely make out even rudimentary spots, yet there are slight traces. I predicted that the young of this species, which in the adult has no spots, might show traces of spots. This is a good demonstration of my theory that the ancestral pigeon had elongated pointed spots. Another striking proof is to be found in the coverts and scapulars of the adult bird, where we find, on lifting the feathers, distinctly outlined impressions, corre- sponding in shape and position with reduced spots, but from which the black pigment has disappeared. These vestigial outlines, structural^' defined, were first noticed in a female bird of a dark shade captured " in Jamaica. The outlines were here more perfect than in lighter birds obtained from Arizona and California. Similar imprints are present in the mourning-dove, and here their identification as marks formerly filled out with black pigment is freed from every shadow of doubt by chequers in all stages of obliteration. They show us that chequers are 11 By Dr. Humphreys. 58 ORTHOGENETIC EVOLUTION IN PIGEONS. not mere blotches of pigment added to an undifferentiated surface, but structu- rally denned areas, which are the skeleton outlines or blank forms once filled out with pigment. That they should persist for a time after the pigment has ceased to appear is not surprising. The process underlying the evolution of color-patterns, so far as illustrated in the species referred to, only takes us back to the chequered condition as a starting- point. How pigeons came into possession of chequers is quite as interesting a question as how they are losing them. This I will briefly indicate, in order to make clear how the same process of evolution reaches far back to the primeval history of the pigeon group and, in fact, to that of the whole class of birds. As before remarked, we find in the turtle-doves of the Old World a pattern which seems to have been the aboriginal pattern for the pigeons and other birds as well. With this pattern as an archetype it is possible to get an orientation of the whole field of avian patterns and to thread our way through what before seemed an impene- trable maze of multifarious variations, with no discoverable beginning or end or order. The pattern is very simple, each feather having a dark center and a light edge. This simple unit or elementary pattern, repeating itself with little or no variation in each feather, results in a uniformly spotted pattern for the bird as a whole. This pattern is well-preserved in the oriental turtle-dove from Japan; but in the adult it has already suffered reduction in the head, neck, and breast regions, only a patch being left on each side of the neck. In the young (in color, pi. 2, Vol. II) the pattern is carried out on these regions, but less strongly than elsewhere, and without any distinct patch on the neck. On the wings and back the juvenal pattern is the same as the adult. The young bird, then, gives no intimation of an earlier pattern, but in striking directly into the general surface-pattern of the adult bears witness to its primeval character. In birds that have made considerable departure from this type the young generally repeat the pattern in their first feathers and exchange it for the later- acquired pattern at the first molt. The adults may retain the original marking in some regions while advancing beyond it in other parts, and in such cases we get the clearest evidence in transitional phases of the direction of modification. Even in the widest departures, where every spot has vanished in the adult plumage, the young bird frequently exhibits more or less perfect traces of the old marking and sometimes requires several molts to reach its mature condition. Without farther digressing into the general field of color-patterns, it remains to explain the genetic connection between the chequered and the spotted type, and to show how both types originated in and have been diversified by the same general process of reduction of pigment. The light apical edge of the feather, by the presence of which the central field becomes defined as a spot, represents a first step in this direction. That this is the correct interpretation is conclusively shown in the turtle-doves and their nearer allies. Taking the oriental turtle-dove as the least modified in this direction, and comparing it with the European species Turtur turtur (color, pi. 2, Vol. II; wing, pi. 22, Chapter V), we see at once that a distinct advance has been made — the light edge is wider and the spot correspondingly reduced, becoming more sharply pointed. Tn another species, the Surate turtle-dove (Spilojielia suratensis), the reduction is carried still farther (pi. 24, Vol. II), having only a mesial streak. In THE ORIGIN AND RELATIONSHIP OF THE ROCK-PIGEONS. 59 the Chinese turtle-dove (Sp. chinensis) the spots have been completely obliterated (pi. 23, fig. A). The neck-spots have been carried to higher and higher stages of differentiation in these four species, and in the order mentioned, 12 so that there can be no doubt as to the direction of modification. The large wood-pigeon (Columba palumbus) of Europe has departed still more widely (text-fig. 6, Vol. II) and evidently from the same type, having lost all its black spots, except a few in the neck patches, which have retreated so far from the tips of the feathers as to be concealed. The gray plumage and the white streak along the edge of the wing mark a plane in the evolution of this bird very nearly identical with that of the white-winged pigeon. A little higher plane has been reached by our band-tailed pigeon (Columba fasciata) of the Pacific coast, which is also a species of turtle-dove derivation, as shown in the neck-marking and in the voice and behavior (text-fig. 5, Vol. II). The ring-doves forming the large genus Streptopelia, embracing thirteen 13 species, are a branch arising directly from turtle-dove ancestors. They have all lost their dark spots everywhere except in the black collar, and the only remaining visible trace of the spots elsewhere is the narrow light edge shown at the tip of each feather in the juvenal plumage (pi. 8, Vol. II). The little red ring-dove (Streptopelia Jmmilis) stands nearest to the turtle- doves, both in color (pi. 21, Vol. II) and behavior. In this species the red is but an extension and intensification of the light edge of the turtle-dove feather. Lifting the scapulars and inner secondaries, we still find obsolete remnants of the dark centers, which have been so reduced as to be wholly concealed by the over- lying feathers. The shaft of the feather still remains dark, while in the common cage ring-dove even this reminiscence of the dark spot has been bleached. In the ring-dove species the reduction process has not only washed out the spots, but also advanced from the darker brown-red to gray, isabelline, and blond, and finally to pure white. Albinism is the finishing stage of the process we have been following. In the common domestic doves it advances through paler and paler shades of gray, the bars growing always narrower. The white rump in the rock-pigeon has followed the pale-gray rump, and this the darker gray. The white- edged wing seen in many wild species has succeeded the gray-edged wing, and this sequence is still often shown in passing from juvenal to adult plumage. The spotted type originated in a marginal reduction of pigment, beginning at the apex of the feather and extending along both sides. Further reduction advanced in the same direction, and so evenly on both sides as to leave the longitudinal axis of the spot generally coinciding with the shaft of the feather. The main advance of the process as a whole was antero-posterior in direction, while in the individual feather it was primarily and predominantly in a postero-anterior direction — i.e., from the apex towards the base. The opposition in direction is apparent rather than real; the free end of the feather is first in development, and hence the direction of reduction is here the same as that of development, as it is in the case of the body as a whole. The chequered pattern of the rock-pigeon originated from the spotted type, by division of the single central spot into two lateral spots. The general course of 12 These neck-murks are fully illustrated and their behavior in hybridization is fully described in Chapter XVII, Vol. II.— Ed. u Seventeen species were later recognized; see Chapter XV, Vol. II. — Ed. 60 ORTHOGENETIC EVOLUTION IN PIGEONS. reduction is the same for both types, but a slight difference in detail is superadded for the chequered type. The light edge, which is an important and permanent feature in the spotted type, is very narrow and transitory in the chequered type, disappearing with the molt of the juvenal feathers. Within this narrow light (pale brown) edge appears a zone of gray that usually enlarges at the middle and extends inward in the form of a wedge, or as a mesial streak, dividing the dark spot more or less deeply into two lateral chequers with pointed tips. The gray area, which arises and enlarges at the expense of the black spot, varies greatly in form and extent in different feathers and in different indi- viduals. It may be wholly absent in some feathers, the black spot remaining undi- vided, as in the turtle-dove; it may take the form of a narrow crescent more or less thickened at the middle; it may have the form of a triangular spot, as in the guinea- pigeon (Columba guinea) of Africa, 14 or be more obtusely pointed, as in the spotted pigeon (Columba maculosa) of South America (pi. 18); it may expand over the whole width of the feather and enlarge mesially so as to deeply or completely divide the black; the mesial streak may expand equally on both sides of the shaft until the two spots are left as edge streaks, or unequally, leaving more or less of one spot than of the other. The mesial extension of the reduction process, which prevails in the chequered type and is its chief peculiarity, is not after all an exclusive distinction. A light mesial streak dividing a central dark spot may and actually does occur in the breast feathers of some of the young of the European turtle-dove; and thus we see that both primary characters — the single central spot and the two bilateral spots — may occur together in each type. I have seen the same mesial streak in the juvenal upper wing-coverts of Leptop- tila (pi. 43). It occurs also in the young robin (see pis. 56, 57), where it closely simulates conditions presented in domestic pigeons. It is a permanent condition in a large number of birds, the female jungle-fowl being an example. We have seen that the process of reduction tends to sweep the whole surface and in the same general direction. The conclusion supported by comparative study admits of experimental confirmation. We may take pigeons of the two-barred type, and try to advance from this condition to that of the chequered type, by selecting in each generation birds with the widest bars, and especially any that may have a trace of a third bar. This I have tried 15 continuously for six years and with several different stocks. I have not been able to establish a third bar, or to extend chequers in front of the vestigial third bar, which is often found. With pure-bred birds, not allowed to mingle with chequered birds, I believe it is impossible to advance from bars to the chequered state. With chequered pigeons, on theother hand, it is fairly easy to advance in the opposite direction, gradually clearing the field and leaving two bars. The process has been carried to the point of completely eliminating the bars. From these considerations we can readily understand why the stock-dove, 16 which has, at least in many cases, a vestigial third bar, quite like that in domestic 14 See Chapter IX and illustrations. The present discussion of the "guinea-mark" — or triangular white tip of the feather — may serve as an introduction to the "guinea-mark" mutations, which are described and illustrated in Chapter IX. It is partly for this reason that some of the illustrations dialing with this character are placed with the references to them in the present chapter. — El). 15 As noted in Chapter II. — Ed. '• Columba cetiax, discussed and figured in Chapter III. — Ed. PLATE 18 Adult Bpotted pigeon, Columba maculosa. Photographed July 21, 1910; colored by Hayashi. 0.66. This South American species shows the "guinea-marks" ( triangular white tip of wing-coverts) in a form slightly modified from that found in the African species, C. guinea, with which it should be compared (pi. 76) THE ORIGIN AND RELATIONSHIP OF THE ROCK-PIGEONS. 61 pigeons, never appears in chequered dress. It is moving in the other direction, and no reversal of course is now open to it. Moreover, the various conditions of chequers and bars in rock-pigeons and domestic races are self-explanatory from the same point of view. We could not explain how two bars could arise de novo in a clear gray wing- surface; but Ave can see how a sweeping reduction process, antero-posterior in direction, would leave two or more rows of chequers cut to dimensions that would coalesce in transverse bars at the posterior end of the wing. The great variations in the width of these bars, the serrate edge of the wider. bars, the more eventy cut edge in narrow bars, the presence of one or two rudimentary bars, wholly concealed from view, in short, all the peculiarities of the wing-pattern, become intelligible as soon as we discover the nature and drift of variation. To attempt to explain all this as the work of natural selection would lead into an endless tangle of conjecture that would leave even the simplest facts as unap- proachable mysteries. Natural selection has probably had much to do with the end-stages in the evolution of characters, but little or no direct influence in originating them. The two-barred condition has been reached in the simplest possible way, not by accidental variation of chance mutation, but by progressive modification of a chequered condition previously established. The long coverts and the secondaries, which are the larger of the chequered feathers, have the larger spots. In the course of a progressive reduction of pigment, affecting all parts of the chequered area, but advancing from before backward, the spots in the lesser and median coverts will reach the point of disappearance at a time when the spots on the long coverts and secondaries have only been cut down to a half or a third of the initial size. At that time we should have left these two rows of spots, reduced in number at the lower ends, where they are weak at the start, and shortened up by a more or less even cutting off of their ends to square or roundish chequers. The two bars, as wholes, would of course have the position and curvature prescribed in the arrangement of the feathers bearing them. Standing alone on a pale-gray ground, these bars would gain immensely in conspicuity and utility as ornamental recognition-marks. The advantage of all this to the species, whatever it be, would be merely an accident of the situation presented at this particular point in a progressive series of modifications. It is conceivable that the utility of the bars might be great enough to give natural selection a chance to step in and bar the way to further reduction. But the process of obliteration has certainly gone much farther in many other species. There may be stages in the process which suggest utility; but when we consider the whole series of stages and note that the process runs on, sweeping away the stages which we imagine to be most useful, we are left with the conviction that some general principle underlying the course of events has not yet been fathomed. The mutationist is compelled to take his stand on immutable unit-characters. A character may fluctuate to and fro, but it never loses its balance, except by a sudden transformation that makes it a wholly new character and its bearer a new species. To the mutationist the bars of the Columba livia, both as a whole and in their component elements, would be viewed as essentially fixed units. As we look at the character in the mature state, it seems a perfect picture of immutability. To 62 OKTHOGENETIC EVOLUTION IN PIGEONS. reach such a definitely localized and apparently rigid character, it might seem that evolution had come to a halt, that the character was caught in a mutation-trap, from which there could be no escape. But even while we are looking at this picture the selfsame process that produced it may be preparing another picture to appear after a molt in more or less different guise. Can a single individual mutate from plumage to plumage, i.e., become a new species? When we look around among allied species and see these same bars reduced to about half dimensions in the rock-pigeon of Manchuria {Columba rupestris, text- fig. 23), reduced to mere remnants of two to six spots in the stock-dove {Columba oetias), carried to complete obsoletion or to a few shadowy reminiscences in the secondaries in Columba rufina of Brazil (pi. 46), gone past return in some of our domestic breeds and in many of the wild Columba? 17 — when we see all these stages multiplied and varied through some 400 to 500 wild species and 100 to 200 domestic breeds, and in general tending to the same goal, we begin to realize that they are not to be regarded as permanent halts, but rather as slowly passing phases in the progress of an orthogenetic process of evolution, which seems to have no fixed goal this side of an immaculate monochrome — possibly none short of complete albinism. Even in cases where natural selection has probably played a conspicuous part in modifying and beautifying these marks, e.g., in the crested pigeon of Australia (Ocyphaps lophotes), we find that the reducing process has not been brought to a standstill. Indeed, a careful comparison of the juvenal and adult plumages in both sexes shows that differentiation has been gradual and continuous and that it is still in progress in the bar of the long coverts, the homologue of the anterior bar in the rock-pigeons. What we see now going on in this bar has been already achieved in the white-bellied plumed dove 18 (Lophophaps leucogaster) of the same country, and is now progressing down the next and last bar. 19 A most striking demonstration of progressive orthogenetic differentiation, still advancing and even cutting through the brilliant coloring, which, in part, we attribute to natural selection. SUMMARY. 20 The wild rock-pigeons present two very distinct color-patterns: (1) the chequered type and (2) the barred type. Two black wing-bars on a gray ground have always been held to be the more primitive pattern, and birds of this pattern are supposed to represent the typical Columba livia. The form with black chequers evenly distributed over the wing and back, although once named Columba affinis, as a distinct species, was regarded by Darwin as a variety derived from the two-barred rock, and his opinion has stood undisputed. It appears from a comparative study of many species of wild pigeons, and from a study of the variations in domestic species, that the relationship is just the reverse: C. affinis is the original rock-dove and C. livia is the derived type. Domestic pigeons have come from both sources. " Illustrations of some of these are given in Chapter VII. — Ed. 18 A good picture of this dove was given by Mr. D. Seth-Smith, in the Avicultural Magazine, December 1906. (The wing-bars of this bird are here shown in text-fig. 9. — Ed.) 19 A detailed account of this case is now in preparation (the account was not completed. — Ed.). 20 The abstract of an address of 1903 (Biological Bulletin, 1904, Vol. VI, p. 307) is used as a summary for this chapter. The four final paragraphs have been substituted by the editor, from the manuscripts for the completed chapter, for the two final paragraphs of the author's abstracts. THE ORIGIN AND RELATIONSHIP OF THE ROCK-PIGEONS. 63 Columba qffinis, however, is not the most primitive form among the wild pigeons. 21 It was derived from a more ancient type, best preserved in the turtle-doves (Turtur orientalis and T. turtur). In the turtle-dove type each feather has a dark center and light edge. The turtle-dove pattern is at the same time a general avian pattern. The turtle type and the rock type coexist in some forms (e.g., Phaps chalcoptera) . The two-barred pattern of Columba livia was reached in the simplest way by an even reduction of the dark pigment, which would result, at one stage, in leaving remnant spots on the long coverts and the secondaries. The process of reduction has run in one direction in many, if not all, pigeons, and present species have reached different stages, varying all the way from a uni- formly spotted condition to four, three, two, one, or part of one bar, or no bars. The light apical edge of the feather represents a first step in this direction (reduction). The spotted type (many species) originated in a marginal reduction of pigment, beginning at the apex of the feather and extending along both sides. Further reduction advanced in the same direction, and so evenly on both sides as to leave the longitudinal axis of the spot generally coinciding with the shaft of the feather. The main advance of the process as a whole was anterior-posterior in direction. The chequered pattern of the rock-pigeon originated from the (earlier) spotted type by division of the single central spot into two lateral spots. The general course of reduction is the same for both types, but a slight difference in detail is superadded for the chequered type. The light edge, which is an important and per- manent feature in the spotted type, is very narrow and transitory in the chequered type, disappearing with the molt of the juvenal feathers. Within this narrow light (pale-brown) edge appears a zone of gray that usually enlarges at the middle and extends inward in the form of a wedge or as a mesial streak, dividing the dark spot more or less deeply into two lateral chequers, with pointed tips. The gray area, which arises and enlarges at the expense of the black spot, varies greatly in form and extent. It may be wholly absent in some feathers; it may take the form of a narrow crescent more or less thickened at the middle; it may have the form of a triangular spot as in the guinea-pigeon (C. guinea) of Africa, or be more obtusely pointed as in the spotted pigeon (C. maculosa) of South America. 21 The succeeding chapters on the Turtle-Dove Pattern in the Phylogeny of Pigeons furnish much of the evidence for this and the several following conclusions. — Ed. CHAPTER V. THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 1 But the causes and conditions of variation have yet to be thoroughly explored; and the importance of natural selection will not be impaired, even if further inquiries should prove that variability is definite and is determined in certain directions rather than in others, by conditions inherent in that which varies. It is quite conceivable that every species tends to produce varieties of a limited number and kind and that the effect of natural selection is to favor the development of some of these, while it opposes the development of others along their predetermined lines of modification. (Huxley, Darwiniana, p. 223.) GENERAL STATEMENT. Color and color-marks present a wide range of the most puzzling problems— the shades of the ground-color, the endless diversity of specific color-marks, the distribution of color and color-marks geographically, their age sequences, regional differentiations, form metamorphoses, sexual correlations, seasonal changes, pro- tective adaptations, irregularity, lawlessness, non-amenability, non-conformity to laws of heredity. What, at first thought, could appear more insignificant than a simple color-mark of a feather? Can such lifeless surface-characters leatl us directly to central problems in evolution — problems which involve at once variability, adaptation, heredity, the genesis of species, and progressive development? Is not variation here too profuse, too multifarious, too lawless to serve as safe ground for fruitful comparative studies and reliable experimental tests? Such, very likely, might be one's first impressions. As we inquire into the subject more closely, we shall see room for a more favorable judgment. Where and how shall we study variation, if not in that which exhibits it abun- dantly? Great variation in outward expression implies a correspondingly high grade of plasticity in the internal foundations of such manifestation. Whether we imagine that each mode and degree of expression has its special primordium, or prefer to refer all the various expressions to one common germ-substance, it is obvious enough that in these color-marks and the patterns which they form we have a wide-ranging variability to deal with, with endless opportunity for both natural and artificial selection and for experimentation. If the very exuberance that confronts one in the feathered world be somewhat appalling on first approach, it costs but little effort to see that it all counts in favor of the investigator. The greater the variability, the closer, in general, will be the connections between stages, and the easier it will be to catch the trend of phyletic derivation and to discover the common points of departure for whole groups of related color-patterns, and possibly to reduce these points to a single point of depart- ure for the whole bird kingdom. Such a vista, once opened, would orient the whole 1 The arrangement of the several "groups" <>l materials of the chapters dealing with this subject is that of the editor. Little of these materials was put into finished form by the author, and it has not been thought advisable to make extensive changes in the record as found. The amount of textual matter is, at, many points, more limited than is desirable. The statements here given — supplemented by the several very full considerations of previous chapters — will, however, doubtless prove sufficient to guide the reader through the great number of illustrations. Further, the legends accompanying the plates are often so full and complete as to render text less essential. The plates and figures, if given alone, would go far toward accomplishing the aim of these chapters, namely, to show that the most divergent forms — of pigeons, and of still other groups of birds —exhibit one or another stage or degree of transformation of an identified ancestral color-pattern. — Ed. 64 THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 65 field, disclose the direction of evolution, provide the investigator with a key to the natural order of sequence in color-patterns, enable him to detect and to demon- strate orthogenetic evolution, if such there be, and to discriminate nicely between this and the results to be ascribed to natural selection and other intervening factors. From such a vantage-ground one may proceed to work out details of evolutional progress by the aid of comparative study of patterns at all stages and ages of development. Juvenile phases of color-patterns become luminous as recapitula- tions in the sense of the biogenetic law and do not stand as isolated prodigies of natural selection or as meaningless exhibitions of mutations. 2 In the 500 or more perfectly distinct wild species (see table 1) with phyletic relationships of easy determination in most cases; in the 200 or more domestic races, all springing primarily from a single known wild stock 3 — in all this abundance of natural and cultivated forms, with a multitude of comparatively simple charac- ters, the evolutional history of which can be largely deciphered in many cases through comparative and experimental study, we find inexhaustible material for just such test-cases as we desire. The more generally distributed patterns, such as occur only in the first plumage, or only in this and the adult female plumage, furnish the chief problems. The patterns of first interest are: (1) the light-edged feather; (2) the dark-edged feather; (3) the dark-centered feather; (4} the transversely barred feather; marginal bar alone dark and light, or light and dark repeated, V-shaped, vermiculate, wavy; (5) the two-dotted feather; (6) the cross-barred wing; (7) the relation of black and iridescence. The principles found to obtain in plumage patterns are: (1) The juvenal plumage presents the earlier type. (2) Females are less modified than males. (3) The upper surface is more modified than the lower surface. (4) Repetitive marks (cross-bars) are multiplied from the tip inward toward the base, the tip bars being the stronger and best defined. (5) The tendency is sometimes (usually) to lose marks and gravitate to uni-color or whole-color; other times to strengthen and extend (?) the marks as ornamental recognitional, warning marks. (6) The pale edge of the first feathers is partly directly continuous with the down, and perhaps this relation accounts for its pale color (like the down). The tips of the barbs come nearest to the primitive color, because first formed. The pale edge varies in width — is wide in the turtle-doves, narrow in Columba livia, and obsolescent in young inca and others. (7) The more primitive the type of color the quicker it is reached; e.g., the Japanese turtle-dove reaches it in the first plumage, while in other families the type of the species is not reached until the adult plumage appears. Again, the inca-dove is more primitive than the geopelias, and it is losing the light edge, which is so con- spicuous in geopelias. 2 These remarkable phenomena, which, as a rule, in normal development, furnish only fragmentary and discon- nected parts of past evolutional history, maybe so extended bysuitable experiments (as noted elsewhere. -Ed.) as to demonstrate complete continuity of stages in the passage to the adult pattern. J It is a great mistake to resort exclusively to domestic races, for here the ancestry contains so many unknown elements that it is often impossible to refer phenomena to their proper sources. Even the so-called "pure-breeds" are decidedly impure as compared with pure wild species. The ideal situation, as regards material, is to have pure wild species in abundance as the chief reliance and allied domestic races for subsidiary purposes. 66 ORTHOGENETIC EVOLUTION IN PIGEONS. (8) The inure the type is modified the longer it takes to reach it. This is shown in highly bred domestic pigeons and in the males of many species of birds, where to reach the much-modified adult type several molts are sometimes necessary. The biogenetic law certainly holds, and often in the most striking manner, in the sequence of color-patterns in individual development. THE TURTLE-DOVE PATTERN IN THE PERISTERID/E (SEE TABLES I AND 2). The TuuTuniNiE. The pat terns of greatest interest are : (1) neck-marks; (2) dark centers and pale tips. Attention may first be directed to the neck-marks: In the turtle-doves of Europe (Turtur turtur) and those of Japan (T. oricntalis) there are to be seen in the first feathers no sign of the adult neck-mark, except that 3 to 5 rows are better defined as rows than the rest (see pi. 36 of Vol. II). 4 This much is precisely the case in the Table 1. — The Five Families of Columboe according to Count Salvadm i .' Order COLUMB.E (5 families). I. Treronidse. II. Columbidae. III. Peristeridae. IV. Gouridae. V. Didunculidae. Asia, Malay Archipel- ago, Japan, Africa, Philippines, Austra- lia, mainly tropical and green plumage and many fruit-eat- ing. Tail, 14 feath- ers (12 in 2 genera; 16 in 1 genus). Columba is universal. Tail, 12 feathers. Tarsus generally shorter than middle toe. Large, mostly tree-pigeons. Widely distributed. Many genera and species. Mourning-dove, ground-doves, geopelias, geotry- gons, Nicobar crested pigeons, turtle-doves. Tail, 12 feathers mostly; geopelias, 14 or 12 feath- ers; Zenaidura, 14; bronze-wing, 16;otidiphaps, 20 feathers. Tarsus equal to or longer than middle toe. Smaller, many ground-doves. Papuan Islands; 6 species; head cresl ; tail feathers, 10. Samoa Island; 1 species. Hill hooked. Low- er mandible toothed or serrated. Isolated and distinct from true pigeons. 1 genus. 6 species. 1 genus. 1 species. Subfamilies under Families I, II, and III. Family I. Family II. Family III. 1. Treroninae. 2. Ptilopodinee. 3. Carpophaginse. 1. Columbine. 2. Macropygiinee. 3. Ectopistinee. 1. Zenaidinse. 5. Phabinae. 2. Turturina?. 6. Geotrygoninse. 3. Geopelinae. 7. C'aloenadinae. 4. Peristerinae. !1. Treroninae 7 genera. 43 species. 2. Ptilopodinae 17 genera, 100 species. 3. Carpophaginae 12 genera, 61 species. 3 subfamilies 36 genera, 204 species. il. Columbinae 4 genera, 65 species. 2. Macropygiinee 4 genera, 33 species. 3. Ectopistinee 1 genus, 1 specieB. 3 subfamilies 9 genera, 99 species. II. Zenaidinee 4 genera, 13 species. 2. Turturina? 5 genera, 28 species. 3. Geopelinae 3 genera. 8 species. III. Peristeridae J 4 - Peristerinae 6 genera, 14 species. 5. Phabinae 12 genera, 25 species. 6. Geotrygoninae 9 genera, 69 species. 7. Calcenadinae 1 genus, 2 species 7 subfamilies 40 genera, 159 species. IV. Goviridae 1 genus, 6 species. V. Didunculidae 1 genus. 1 species Total 5 families, 13 subfamilies, 87 genera, 469 species. 1 For the Juvenal and adult neck-marks of various grade3 of risoria X turtur hybrids see plates 35 to 39 and text-fig. 11 of Vol. II.— -Ed. Catalogue of Birds in the British Museum. Vol. XXI. 1893 THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 67 Table 2.- — The family Prrislcridae according to Count Salvadori.' Family and subfamilies. 7 Plumage distinctions. Genera. Habitat. Family PERISTERIDjE. b. Sides of neck with metallic gloss. c. Scapulars and wing-coverts spotted, except in Melopelia. d. Tail, 14 feathers (12 in one). a. Side-neck spots or dark collar. c. Upper feathers fulvous or pale- edged; dark centers or plain. d. Tail, 12 feathers a. Neck with bars — not metallic. b. Upper feathers black-edged in most species. c. Tail, 14 feathers (or 12) a. Metallic spots on wing (ex- cept Metriopelia melanop- tera). b. Tail, 12 feathers.. 2. Zenaida 3. Nesopelia 4. Melopelia 1. Turtur 2. Homopelia 4. Spilopelia 5. Stigmatopelia 1. Geopelia 2. Scardafella. . . 3. Gymnopelia.. . . 1. Columbula 2. Chamsepelia. . . . [3. Uropelia J4. Oxypelia |5. Peristera [6. Metriopelia. . . . 1. CEna North America to Central America and West Indies. Yucatan and Florida Keys to South America. Galapagos Islands. Southern United States, Mexico, Lower California, West Indies. Europe, Asia, Japan. Africa, India Madagascar region. Africa, India, Oriental Malay Archipelago. Oriental Malay Archipelago. Africa, Turkey, West Asia to India. Burma to Australia. Southern United States, Mexico, Central America, and Brazil. Peru, Bolivia, and Chile. South America. Southern United States, Mexico, ( Central America to South America. Brazil and Bolivia. Brazil. Mexico to Peru and Paraguay South America. South America. Ethiopian region. Africa. India and Australia. Africa. New Guinea. Australia. Do. Do. Do. Do. Do. Ethiopian region. Texas to South America. Ecuador. Mexico to South America. Malay Archipelago and Polynesia Australia. New Guinea. Do. Cuba and Florida Keys. Malay Archipelago and Pelew Islands. (New World and Gala- pagoB.) (Old World.) (India, Malay Archipel- ago, Australia, Southern United States, Central America to Brazil.) (Confined to America.) a. Metallic spots on wing, trans- verse bars in some. b. Tail, 12, 14, or 16 feathers . . . a. No spots on wing (a violet patch on bend of wing in some). b. Tail 12, 14, 20 feathers [a. Neck feathers hackled; very long. lb. Plumage metallic green. [c. Tail white. 2. Tynipanistria. . 3. Chaleopelia 4. Chalcophaps . . 5. Calopelia 6. Henicophaps. . . (Africa, India, Malay Archipelago, Australia.) 8. Histriophaps. . . 9. Petrophassa. . . . 10. Geophaps 11. Lophophaps. . . 12. Ocyphaps 1. Haplopelia 2. Leptoptila 3. Osculatia 4. Geotrygon 5. Phlogienas . . . 6. Leucosarcia. . . . 7. Eutrygon 8. Otidiphaps . (Africa, America, Malay Archipelago, and Aus- tralia.) (Malay Archipelago and Pelew Islands.) s Loc. cit., p. 372. This can not be a correct arrangement of the subfamilies. common pigeon, indicating common parentage. In blond ring-doves (Streptopelia risoria), which have advanced beyond the turtle-dove, we find in all the .young these rows not only better defined as rows in the first plumage, but in more than half the individuals there are plain lateral spots (see pi. 19, fig. B), generally extend- ing over 3 rings, sometimes 4, at other times only 2. In many young, however, the 68 ORTHOGENETIC EVOLUTION IN PIGEONS. spots are absent in the first plumage, or so nearly so as not to be visible to surface examination. 8 In the related red ring (St. hnmilis) all of the young have these spots in the first plumage. 9 This species has more pigment, hence enough to appear in the first feathers. The Japanese ring (St. douraca) is darker than the common blond ring, and therefore more often — almost invariably — has spots in the first feathers. The Damar ring-dove (St. damarensis) is another form more pigmented than the blond ring; in this dove (pi. 19, fig. B) the mark is quite strong in the juvenal feathers and is plainly lateral. In these neck-marks of the collared ring-doves we have, therefore, a good ex- ample of a character which appears earlier in the ontogeny than in the phylogeny — a tendency to appear earlier — what first came in second feathers now comes in the first feathers. 10 The neck-marks of the five groups into which Salvadori 11 divided the former genus Turtur were somewhat inaccurately or inadequately described. These groups were designated: (1) Turtur, (2) Homopelia, (3) Streptopelia, (4) Spilopelia, (5) Stigmatopelia. Groups (4) and (5) are stated (p. 396) to have bifurcated neck- feathers. Spilopelia: ''Feathers of the hind neck bifurcated, black, with two ichite apical spots." Stigmatopelia: ''Feathers of the fore neck bifurcated, black, with two rufous apical spots." It is strange that Salvadori should entirely overlook this feature in Homopelia. Mr. T. H. Newman 12 calls attention to this oversight, and also refers to Temminck 13 as having given a separate figure of one of the bifurcated feathers. These homopelias are very interesting cases, as they have bifurcated neck- feathers, sometimes on the sides and back of the neck, and sometimes on both sides and front of the neck. The latter seems to me to be a departure from the rule that such marks are limited to the side and back of the neck. 8 The black ring or collar in the blond ring (St. risoria) is sometimes narrower at its middle, on the back of the neck, than at the ends. This fact, together with the complete interruption of the ring on the back of the neck in the first plumage, by which it is broken into two portions— one on each side of the neck, beginning a little below and behind the ear-coverts, and growing narrower backward— suggests that the half-ring has arisen by the extension of two spots like those seen in the turtle-dove. This ring is creamy white in the white ring (St. alba). It is often reduced to a mere shadow, or wholly absent, in the first plumage (see Chapter XVII, and plate 31, Vol. II). 9 For the juvenal neck-marks of the risoria X humilis hybrids, see text-fig. 9 of Vol. II.— Ed. 10 Darwin (Descent of Man, Vol. II, p. 153) points out the fact, and makes frequent use of it, that "variations occurring at different ages are inherited at the same age." And, further: "Variations which occur late in life are commonly transmitted to the same sex in which they first appeared; while variations occurring early in life are apt to be transmitted to both sexes; not that all the cases of sexually-limited transmission can thus be accounted for." From the standpoint of the present, all inheritable variations are in the germ. The order in which they come to sight is the order in which the germ has been modified — not the order in which a new unit-character was added, but the order in which the germ as a whole was modified. Darwin's reasoning on the "lopping of the comb" in the hen of Spanish fowl (pp. 151-152) is characteristic of his viewpoint. My idea is that the difference here between the sexes is merely one of strength. Darwin says, too (p. 148), that "characters which are limited in their development to one sex are always Intuit in the other." He had elsewhere written (Origin of Species, p. 12): "A much more important rule, which I think may be trusted, is that at whatever period of life a peculiarity first appears, it tends to reappear in the offspring :it a corresponding age, though sometimes earlier. In many cases this could not be otherwise; thus the inherited pecu- liarities in the horns of cattle appear only in the offspring when nearly mature ; peculiarities in the silkworm are known to appear at the corresponding caterpillar or cocoon stage. But the hereditary diseases and some other fads make me believe that the rule has a wider extension, and that, when there is no apparent reason why a peculiarity should appear at any particular age, yet that it does tend to appear in the offspring at the same period at which it first appeared in the parent. I believe this rule to be of the highest importance in explaining the laws of embryology." " ( 'atalogue of birds in the British Museum, Vol. XXI, 1893. I? Avicult. Mag., Jan. 1908, p. 80. " Hist. Nat. Gen. des Pig. PI. Col. 242, 1823. PLATE 19 B Juvenal, 6 weeks old, Damar ring-dove, St. darmrenm. Shows strong neck- mark in first plumage. The neck-mark is (as In the adult turtle-dove) a sidi spot, not a half collar as in the adult Damar ring. Juvenal blond ring, St. risoria, to compare with above. The first feathers slum weaker neck-marks than the Damar ring. The mark in the Mend ring (v present) is also a sick spot instead of a half collar. The turtle-doves show no color in these feathers in the first plumage. THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 69 Newman figures the painted dove (Homopelia picturatus), commonly called the Mauritian turtle-dove, and proposes to call it the "Madagascar turtle-dove," as Mada- gascar is much the larger island and is perhaps the only true habitat of the species. I have a number of the Senegal turtle-doves (Stigmatopelia senegalensis) , and find that some of them have the whole front of the neck with bifurcated feathers; but one or two individuals show a narrow median front line, 14 in which the feathers are neither marked with black nor bifurcated. The Madagascar birds would be the more primitive type, while the Senegal birds would be the later form. These birds fly without noise, like young doves generally. The dark centers — already described as an early or primitive avian character — most nearly fill out the feather and are least pointed in the Japanese turtle-dove {Turtur orientalis) (see pis. 1 and 2, Vol. II; and text-fig. 3, Vol. I). The position and extent of the black in the wing-coverts is best seen in a series of the separately drawn feathers (pi. 20). The black center blends with the reddish edge or feather tip; i.e., at the boundary of the black we see an intermixture of red, with no sharp separation. The lowermost coverts of the wing show the red replaced by gray. This black center can be regarded as consisting of an upper (inner) half and a lower (outer) half. On the upper tertials the upper half may fill more of the surface than the lower half; in other words, there is an asymmetry such as we see in the black spots of Columba. In the young the centers of the coverts, scapulars, tertials, etc., are sometimes blackish, sometimes dark brown, as in some of the tertials. In the adult the spot is clearer black and more decidedly angular at the tip. 15 In the European turtle-dove {Turtur turtur) the black center is more reduced and more pointed. This is made clear by a comparison of corresponding wing- coverts in the two forms (pis. 20 and 21). The drawings of the entire wings {turtur, pi. 22, fig. 1; orientalis, text-fig. 3) show the same distinction, and that of turtur further shows that it is the anterior part of the wing that has suffered relatively greatest reduction of the black. The pale tips and lateral edges are characters of much interest in the young of turtle-doves. Old types reach adult conditions in their first feathers; e.g., the turtle- dove gets its characteristic light edges in the juvenal plumage. 16 The first young Japanese turtle-dove that I examined seemed to have its entire adult pattern in the first feathers. I have now seen other young as well as many young European turtles, and I find that it is necessary to distinguish in the first feathers the tips and the lateral edges. The tips (apical edges) in Japanese turtles have a decidedly paler color than the lateral edge, the lateral edge being here more nearly the equivalent of its adult color. This distinction is particularly evident in the European turtle, which is much lighter in color than the Japanese. I find, however, that the distinction also holds, though in a lesser degree, in orientalis. In this species the difference can not be seen on the scapulars, upper long coverts, and "It will later be seen, in the geopelias and other forms, that this midfront line and the parts immediately adjoin- ing it are points that differentiate themselves from the rest of the body plumage. 15 In the young a barred effect on the wing and scapulars is produced, owing to the fact that the creseentic rufous edges of adjacent feathers match so well (reminding of Geopelia humeralis). These rufous bars of course alternate with dark bars, which are broader than the reddish or fulvous bars. This is a very simple type of barring, and pos- sibly the starting-point for bars repeated from the tip towards the base, centiipetally or basipetally. "The blue-jay gets the adult pattern in its first feathers, and even the head-crest is there but little less developed in the young than in the old. 70 ORTHOGENETIC EVOLUTION IN PIGEONS. tertials, all of which have about the same color at the tips and lateral edges as in the adult. But farther down on the body the feather-tips lighten up, more or less, and at least some few of the central barbs of the tip will show the lighter color. The third long covert (counted downward) of Turtur orientalis is compared with the same feather of T. turtur in figures 2 and 3, plate 22. The rufous tip of orien- talis is about 2 mm. wide, and it lightens up gradually, becoming cinnamon-brown only at the very edge. There is no well-defined tip distinct from margins in the particular feather drawn, but in the covert that grew next to it there is also a little of the apex that is pale. The dark center is here much broader than the same feather of turtur, but this center varies in extent and intensity in both species, although the maximum in turtur is less than the minimum in orientalis. The Euro- pean turtle-dove is losing these spots. In orientalis the primaries are also rather widely margined with rufous, and remind strongly of young Ectopistes, which is similarly marked, though to a little less degree. Can we say, then, that the light edge seen in all the pigeons in the first feathers indicates a turtle-dove ancestry? I still think so, although it can not be said that Explanation of Plate 20. The original feather color-pattern, as shown in a female Japanese turtle-dove, Turtur orientalis. Figs. 1 to 18, x 1; figs. 19-24, x 2. Hayashi del., May 1903. All feathers from light side of bird. Figs. 1 to 4. — A series running from before backward in the interscapular region of the back. All have a fulvous margin, duller than on wing-coverts and scapulars. Shafts of feathers dark and whole feather gray, darkening somewhat in distal half (this darkening is quite gradual, scarcely showing so strong a con- centration as the artist represents). Figs. 5 to 8. — A scapular series running from front to rear end. Black and pointed centers here become stronger, and red tips more clearly marked off. Figs. 9 to 16. — Wing-coverts: second row long coverts, counting from behind forward, running from above down- ward, or from inner to outer to below middle of row. Here is the typical pattern for the turtle-dove well defined. The blackish centers pale into brown-gray on Nos. 15 and 16. Fig. 17. — An upper long covert (first row). Fig. 18. — The same from a female hybrid between a male chequered homer and female Japanese turtle. Here the fulvous turns into a dull cinnamon that blends with the gray-brown center, which is not sharply defined in any of the feathers. Fig. 19. — First of second row of neck-spot, anterior end of row. Fig. 20. — Fourth of third row of neck-spot. Fig. 21. — Sixth of fourth row of neck-spot. Fig. 22. — Fifth of fifth row of neck-spot. Fig. 23. — Fourth of sixth row of neck-spot. Fig. 24. — Third of seventh row of neck-spot. The feathers of figures 19 to 24 have tips of gray. The black portion is seen but little if the feathers are drawn close. We have here the same dark center, obtusely pointed, and a gray tip instead of light or fulvous tip; this gray is a higher stage, and white is higher still. Explanation of Plate 21. Feather pattern in the European turtle-dove, Turtur turtur. Figs. 1 to 17 x 1 ; figs. 18 to 23 x 2. Hayashi del., Apr. 1903. Figs. 1 to 6. A series of back feathers (interscapulars). 7 to 10. A series of anterior scapulars. 11 to 16. Second row of long coverts, from above to below middle of row. 17. Upper feather of upper row in neck-spot. 18. Anterior feather of upper row in neck-spot. 19. Third feather of second row in neck-spot. 20. Fourth feather of third row in neck-spot. 21. Sixth feather of fourth row in nock-spot. 22. Fourth feather of fifth row in neck-spot. 23. From mid-back of second row in neck-spot. In general this type stands above the Japanese turtle in (1) a reduction in area of the black center and in giving it a sharper definition; and (2) in the whitish or very pale gray edges of the feathers of the neck-spot. PLATE 20 ■ ■ I m- : £&. =«b .. ^5gg§£ id - . Feather panern in the Enropean turtle-dove, Turtvr furfur. Pige. 1 to 17 < 1; Hgj Hayashi del.. Apr. 1903. THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 71 this light edge is homologous with the whole fulvous margin. This margin has been widening and has now become much wider than it was originally. 17 In the young of turtxir, the dark center is reduced — i.e., it retreats farther from the feather margin (fig. 3, pi. 22) — and here the rufous lateral edges are wider than in orientalis, the rufous seeming to take the place of the retreating black. In these feathers the tip is cinnamon-brown, only a shade darker than the down itself, and is 1.3 mm. wide at the center, from which it narrows outward. There is no concentration of dark pigment bordering the proximal side of this tip — such as I see in Streptopelia humilis, Stigmatopelia senegalensis, Streptopelia damarensis (figs. 4 and 5, pi. 22), and other forms that have lost or much reduced the dark center. The third long covert (counted downward) may be said to have the pattern in typical form for this European species. Here the feather's base is gray, becoming distally dark gray, and passes into black in the distal half of the center, the black becoming pointed and reduced to a shaft-line which reaches, but does not extend into, the light tip. The rufous margins are of very considerable width, easily attain- ing 5 mm. The rufous of the margin has a sprinkling of blackish beginning on its inner half and increasing towards the dark center, thus blending with the center, so that we have no sharp boundary. In the breast-feathers of a juvenal turtur (about 2 weeks old), and in the same feather of its nest-mate, I find what I had not previously noticed in other young turtles — a pale, whitish shaft, or a narrow shaft-streak, expanding at the feather's tip into a triangular cinnamon-brown spot, comparable in shape and position with the apical spot seen in the robin and some other forms. (Later, I recognize in this mark of the turtle-dove an indication of the guinea-pigeon's apical mark of white.) The breast-feathers are all tipped with the cinnamon-brown and the median spot forms part of the tip or pale edge; i.e., the color of the spot and the edge are the same and continuous. In other specimens of European turtles I do not recognize this median streak and spot. Neither do I see it in the Japanese turtles. Among the speckled-neck doves (Spilopelia) we find the black center further reduced. In the Surate turtle of Latham (Sp. suratensis) Salvadori (p. 445) notes that the "black contracts to a central streak," and that this reduction is accompa- nied by a modification (in the scapulars and upper back feathers) of the pale edge into "two pale vinous isabelline spots" 18 (pi. 24, Vol. II); in the tiger turtle-dove "The author later noted the neck-roark, dark center and pale tip of orienlalis and turtur in Turtur ferrago. Salvadori's description (pp. 400 and 408) makes it clear that all these are also present in the two remaining species, T. isabellinus and T. lugens of the genus Turtur. — Ed. 18 The Surate turtle of Latham (Spilopelia suratensis), in its juvenal stage, represents, in its color-pattern, quite closely >S>. tigrina. The female does this more closely than the male. I have a pair now (July 1910) in full juvenal plumage. The male shows in its scapulars a decided step towards the adult pattern— most of these feathers showing the "two pale vinous isabelline terminal spots, enlarging and spreading upon each side of the feather" (vide Salvadori, p. 445). The coverts of the whole wing exhibit the "dark mesial stripes" of the tiger turtle. The female shows less of the pale isabelline spots on the back, and its wing-pattern closely resembles that of the tiger turtle. I have preserved three of the posterior scapulars of the right side of the young male in order to show the spots and at the same time to show that these spots are nothing but transverse bars in origin. Every feather in the juvenal stage of all pigeons has a more or less pale-brownish apical edge. This apical bar is followed in the Surate turtle by a dark bar, and this by another pale bar. It is this subterminal second pale bar out of whirl, the pale-isabelline spots arise. The evidence (if this is plainly seen in these plucked juvenal feathers; the idenl ity of the spot with the second pale bar is well shown. The enlargement of the lateral elements of the bar takes place through the suppression of one to several of the subterminal dark bars, the light bars thus coalescing to form the two "terminal spots." The case of the .Surate turtle is not an isolated case. Out of these same transverse bars arise the pair of white dots on each of the wing-coverts in the little diamond-dove (Gcopeha cuneata), as I have found out by studying the juvenal plumage. These transverse bars play a great role in giving rise by modification, fusion, etc., to specific color-types. 6 72 ORTHOGENETIC EVOLUTION IN PIGEONS. (Sp. tigrina) the dark center is reduced to a narrow mid-streak. In still another, the Chinese turtle (Sp. ehi?iensis), it is absent (pi. 23, fig. A). The evolution of the dark center in these forms, then, was as follows : We have to start with a form most nearly represented in Turtur orientalis of China and Japan. The centers of the wing-feathers have the black rounded out full behind, leaving only the apical reddish bar or tip. In Turtur turtur of Europe the dark center becomes reduced, ending in an angu- lar point behind. In Spilopelia suratensis the black is still further reduced — to a dark mesial stripe (and has also bifurcated neck-feathers). In Spilopelia chinensis even the mesial stripe has departed, and the neck-feathers, tipped with white, are bifurcated. In these two last-named species we have the neck-feathers not only bifurcated, but a large number of rows are differentiated. Explanation of Plate 22. 1. Left wing of an adult European turtle-dove, Turtur turtur. Natural size. Hayashi del., Sept. 1898. For comparison with wing of the Japanese turtle (text-fig. 3). (Figs. 2 to 5 x 2/1, Hayashi del., 1903-08). Dark centers smaller in size and more pointed or angular at tip ; the pale edges correspondingly increased in width. These modifications effect their most striking change in the appearance of the wing in its anterior part. 2. Third (inner) longer covert of juvenal (4 weeks) T. orientalis. The pale tip is 2 mm. wide, and the dark center leaves only slender lateral edges. 3. Third (inner) longer covert of juvenal (2 weeks) T. turtur. The pale tip is 1.3 mm. wide; the black rufous lateral margin is 5.5 mm. wide at its widest part. Practically only the shaft is black at the most narrowed posterior extremity of the black center; the black of the shaft reaches to but does not invade the pale tip. A description and comparison of the posterior scapular in orientalis and turtur follows: T. orientalis. — Posterior scapular. The black is more extended and the brown less extended than in T. turtur. The basal part shades into gray. T. turtur. — Posterior scapular. Cinnamon-brown margin, black center arrow-pointed. Basal half gray, grading into black center, which fades into a trace of gray at its border. The third tertials compare as follows: T. orientalis. — Third tertial, i.e., next above longest. The brown in these feathers is darker and richer than in T turtur. T. turtur. — Third tertial. The black spot is arrow-pointed. Posteriorly the brown or dark center fades into gray at the basal half. In both species the pointed black centers mark all of the feathers of the upper surface and wing, being larger and generally less pointed in T. orientalis than in the other species. There is not an indication anywhere that the black center represents a fusion of two spots. 4. Third (inner) longer covert of juvenal St. damarensis. 5. Third (inner) longer covert of juvenal Slig. senegalensis. These four figures show that the dark center and pale edge of the T. orientalis pattern is present in all these species, and is most modified in the less closely related Streptopelia and Stigmatopelia. Explanation of Plate 23. A. Adult Chinese turtle-dove, Spilopelia chinensis. x 0.7. Hayashi del., Dec. 1903. No dark centers in the general plumage, present only in the neck-mark. B. Fifth feather (natural size) third row (right side) of the neck-mark of Chinese turtle (same as fig. A). Shows strong bifurcation, white tip, and dark center. C. Wing of juvenal mourning-dove and ring-dove (female parent — alba-risoria x risoria-alba hybrid) hybrid; hatched May 9, 1899. Natural size. Hayashi del., June 6 ; 1899. 4 weeks old. There are three small pale gray spots visible on the upper long coverts, and a fourth spot on the next tertial below; the latter spot is entirely concealed. On the next row of coverts, at the same level, are three smaller spots, similarly placed at the lower edge of feathers. On one or two of the upper tertials just a trace of a streak may be seen. This picture should be compared with the juvenal Melopelia leucoplera (text-fig. 14). The influence of the ring-dove (chequers wholly lost from the wing I mother does not lead to a suppression of the chequers of Zenaidura, but to a reduction to a stage nearly that of Melopelia. PLATE 22 1. Left wing of an adult European turtle-dove, T. turtur. Natural size. 2. Third (inner) longer covert of juvenal (4 weeks) T. orientals. X 2. 3. Third (inner) longer covert of juvenal (2 weeks) T. turtur. X 2. 4. Third (inner) longer covert of juvenal St. damarensis. X 1. 5. Third (inner) larger covert of juvenal Stig. senegcderms. X 1. Hayashi del., Sept. 1898. iBifiwt PLATE 23 A Adult Chinese turtle-dove, Spilopelia chinerms. X 0.7. Hayashi del.. Dec. 1903. B. Fifth feather (natural size) third row (right side) of the neck-mark of Chinese turtle (same as fig. A). Shows strong bifurcation, white tip, and dark center. C Wing of Juvenal mourning-dove and ring-dove (female pzrent-alba-risoria X nsona-aba hybrid; hybrid ;' hatched May 9, 1899. Natural size. Hayashi del., June 6, 1899. 4 weeks old. THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 73 * Running back to Turtur orientalis, we come to entire neck-feathers and fewer rows — six down to only three or four. An extension of differentiation towards the mid- back of the neck has occurred in the derived forms, the two spots here coalescing plainly for the upper three or four rows. 19 It is clear that there are two modes of reducing the original black or brown centers: First, by the guinea-mark (the white or light-colored wedge-shaped exten- sion of the light apical edge) splitting the center, into paired chequers; second, by reducing the center through an extension of the whole of the pale feather border. It is this second method — or extension of method 20 — that has been effective in the Turturinse. It is in this way that the somewhat narrowed and pointed centers of the European turtle, the narrow mesial stripes of the tiger and Surate turtles, and the dark shafts of the red ring were formed. Evidence for this latter conclusion, from an hitherto unmentioned source, is easily obtained by hybridization experi- ments. Crossing the Japanese turtle with, for example, the blond ring-dove, which has a whole color — no spots except in the collar — we obtain in the hybrids feathers whose indications of dark centers are confined wholly to their shafts, or to their shafts and the immediately adjoining parts. 21 The pale tip or brownish edge of the feather varies in orientalis and turtur in a reverse direction to that of the dark center, the tip being relatively wider and more pronounced in turtur. Although the juvenal pattern in both of these species is identical in the main with the adult pattern, we may note some differences: (1) The fulvous-edged feathers in the young include not only the wings, scapu- lars, back, and tail-coverts (as in the adult), but also the feathers of the neck, breast, and even those of the head, which in the adult are not so edged, but are uniform gray-brown. (2) The fulvous edges are more nearly straight in the juvenal plumage, owing to the form of the feathers. (3) The fulvous edges are, in general, paler in the young than in the adult. The tertials and posterior scapulars are practically the same as in the adult. This light apical edge seems to be universal in birds. Other features of the color-pattern of some of the collared doves (Streptopelia) than the neck-mark, already referred to, indicate the direct derivation of these species from a form like Turtur orientalis. The 13 species 22 have all lost their dark spots everywhere except in the black collar, and the only remaining visible trace of the spots elsewhere is the narrow light edge shown at the tip of each feather in the juvenal plumage (pis. 16 and 17, Vol. II). The reduction process has not only washed out the spots, but has also advanced from the darker brown-red color to gray, isabclline, and blond, and finally to pure white (Streptopelia alba). Albinism is the finishing stage of this reduction process. The little red ring-dove (St. humilis) stands nearest to the turtle-doves both in color (pi. 21, Vol. II) and behavior. In this species the red is but an extension and intensification of the light edge of the turtle-dove feather. In its adult feathers the light or pale edge is wholly wanting ; but in the first feathers we have a neat ly 19 The neck-marks of the several species here mentioned were separately drawn — much enlarged — and are shown and described in Chapter XVII of Vol. II. — Ed. 20 The whole of the pale edge instead of the ontogenetically older part of it — the apical edge — here takes part in the invasion of the black. "This median streak curiously appears in several of my young pouters (of this year — 1910) which have fallen back from a very light gray with white bars to a much darker gray with red bars bordered with a narrow black. 22 Later (see Chapter XV, Vol. II) Whitman recognized 17 species in this genus.- — Ed. 74 ORTHOGENETIC EVOLUTION IN PIGEONS. marked edge — pale yellowish-brown crescents — on all the feathers; these, however, are best marked on the wing. The rest of the middle part of the feather (exposed part) is a uniform brown, darker in the male (pi. 21, Vol. II), lighter in the female. I have frequently noticed further that the young birds, in many of the wing- coverts and scapulars, have the dark pigment more concentrated on the inner edge of the light margin, giving an appearance that reminds of the dark crescents within the pale margins in the young geopelias. 23 Proximally of this dark bar I usually see an irregular but plain light transverse band — a kind of imperfect or incipient bar. Again, the vinous red of the adult seems to be an extension of the reddish edges of the turtle-dove's feathers. Lifting the feathers, we see that their basal portions are dark gray, the dark graj' frequently extending farther on one web than on the other. The gray center has simply been reduced so that the uncovered part of the feather is all vinous red. The juvenile feathers of this species give strong evidence of its turtle-dove an- cestry (pi. 21, Vol. II): (1) The pale yellowish-brown edges of the feathers incline more to red on the primaries. (2) The neck-spots are always present, sometimes quite strong; but they never meet on the back of the neck. (3) The mid-ribs of the feathers (coverts of wing) arc distinctly dark — a final stage in the loss of dark centers, as shown in hybrids of ring-doves and turtle-doves. (4) One feature is particularly significant. On the tertials (upper inner second- aries) and on the third or fourth posterior rows of coverts we see that the lateral edges as well as the tips are pale; but the lateral edges are decidedly more reddish than the tips, quite closely following the plan seen in the Japanese turtle-dove young. The color is here not so strong as in the turtle-dove, and it is not so well delimited towards the central field of the feather. In fact, it extends inward rather diffusely to or near the mid-rib on the lower web more than the upper (inner) web. There is a subterminal concentration of dark pigment following the light edge, as if light and dark alternated in the manner of transverse bars. 24 The probable relationship of several genera of pigeons is indicated in the fol- lowing diagrams: 25 T. orientalis T. turtur St. humilis 'St. douraca 'Spil. tigrina; Spil. suratensis ^\ St. risoria 'St. alba Spil. chinensis Columbidae (e. g., C. palumbus) Derivation of species of TurturinaV' 9 and part of Family lu Combidae. Turtle-dove type / 1 Nesopelia I Zenaidura I Zenaida ' Chamaepelia Columbidae (e. 2 tfl > -t a o 13 * ," -r OJ o> U T3 T3 -- . — • i— « — — +j OS 7-, rC >. >i ff a ad ffl & 03 e3 S QJ ClJ X = GO cd ,_, i , — ~ GJ 5 +3 — cj - • », Z !z O S S - — . /~N w 1— < 1— < a rH T— i 1 " V s_^ ■x .°9 « V 1 o 8 S t> r 8 s 8 I 8 if e — 6SJ a 00 OS a. 0J & 03 Bt - s a - u 1 s 3 — — 08 Bi _ 1— e o <; o i - ■— N tSJ - - -*-- ti 42 -faj -. _ 1 — 1 A c a gj ej m - i n - & p r: T! - -*-• : 7. GO -** - Q :- — a T3 3 - < O H ffl PLATE 25 . . WiL ' m . I ! i ^ 12 11 10 ■' 8 Selected feathers from juvenal mourning-dove, Zenaidura carolinenm. X 2. Hayashi del. , Feb. 1904. WfcetiRr.nBilhnoiT PLATE 26 PQ aa O O I C3 OS 3 2 3 Haya Haya id iC i— i i— ( X X C QQ & a « "3 q ? | i 1 r S a o 1*1 o •a i EH be ~ g '? A - +a o Irt 5 & aj a> 3 - a 9 T3 n4 o o (-. t* ej as £ £ ■ i O O ne lesser covert from about the middle of the oblique streak has a fairly well defined spot. The spots in most of these feathers are remarkably shadow-like, especially on the median coverts and tertials. There is a median buff freckled streak in each of the median and long coverts. There are buff apical marks, and a faint trace of a subapical dark bar. 31 It seems that no special illustration was prepared of the oblique bar in Zcnaidura. — Ed. 80 ORTHOGENETIC EVOLUTION IN PIGEONS. foundations for what is found in Columbula. 32 Moreover, a glance into a wholly different family, the Columbidae, is worth while in a consideration of the status or origin of the oblique wing-bar of C. piqui. In a 7-weeks-old common pigeon I find an oblique light streak, in which the feathers are gray, without any edging except a trace of darker gray. 33 All the other feathers (coverts) have a tinge of brown in the gray, and are narrowly edged with 32 In a Juvenal ground-dove (Chamoepelia) from Santa Martha, United States of Colombia, I find the oblique tract is more strongly marked with spots than other coverts in front of it or behind it. Furthermore, the spots are darker, and the light edge is reduced in width. 33 In another common pigeon of 15 days — from a cross of a red barb and a red and white jacobin — of blackish-brown color I find feathers with dull reddish tips 1 mm. wide. About 2 or 3 rows of pin-feathers of the oblique tract are beginning to unfold, but are as yet concealed by overlapping lesser coverts, except for 3 or 4 somewhat longer ones near the bend of the wing, which protrude 3 mm. to 5 mm. The interesting point here is that the oblique-tract feathers still show the apical mark, but only about half width, and about half depth of color. Explanation of Plate 28. A. Left wing of juvenal female Chamoepelia passerina. x 2. Toda del. Completed October 20, 1903, at age of 12 weeks; but wing was outlined four weeks earlier, at age of 8 weeks, just as a few feathers had been molted. All feathers were finished as first feathers. (1) There are 35 spots visible, some of them weak; (2) posterior areas are much like those in C. affinis; (3) all leathers are light-edged. B. Feathers corresponding to figures of male shown in plate 27. x 2. Toda del., Sept. 1903. Figs. 1 to 4 from juvenal female Chamoepelia passerina (left wing, 7 weeks old). (1) Pigment is weaker than in adult (see pi. 27); (2) the red- vinous of the adult is here pale vinous-gray or gray; (3) in the feathers corresponding to ti to 13 of plate 27 the dark pigment is found on both sides. These juvenal dark spots are more elongate, as in Columba; mid-rib is light fulvous in color (reminds of young Leptoptila) ; (4) all feathers are edged with pale whitish buff; (5) in 14 to 15 the chequers are pointed behind as in ('. livia (chequered type); evidently homologous in the two forms. Fig. 1. Thirteenth feather of plate 27; third row of coverts (middle). Notice two lateral spots or chequers; also median light-buff streak (as in Leptoptila). 2. Fourteenth of plate 27; middle second row of coverts. Only one chequer. 3, 4. Fifteenth of plate 27; two upper (inner) long coverts. The median light streak is very narrow. Chequers are pointed and marginal, except for light edge. C. Selected feathers from a juvenal Chamoepelia passerina. x 3. Hayashi del., 1905. Feathers taken from left wing of mounted skin September 15, 1905, and drawn for transitional phases between central spot and lateral spot or chequer. On the wing we find about 5 rows of spots: First row, 3 inner tertials. Second row, 3 inner long coverts. Third row, 5 inner posterior median coverts; one or two feathers are double-spotted; spots on inner side are rudiments. Fourth row, 4 and 5 inner mid-median coverts; several of these feathers double-spotted, but the inner spots still small. Fifth row, 4 to 6 inner anterior median coverts; all double-spotted, but the inner spots smaller than outer ones. In front of these four rows of coverts come the lesser coverts on front edge of wing. These do not show spots on surface when the feathers are in situ, but several of the posterior row of lesser coverts bear spots. When the lesser coverts are lifted we find that many of them have the turtle-dove central spot, pointed distally, and some of these present beautiful transitions between the turtle spot and the lateral spots, especially in the posterior row (or two rows) of these coverts. The coverts have a whitish edge or tip; those bearing spots and some that have no spots have a mesial rufous streak. This streak is the same as those seen in Zenaida, Geotrygon, European turtle-dove (some young), and in some other species. The streak is sometimes only of the width of the shaft, sometimes it widens a little, extending to differ- ent lengths on the barbs. The feathers of the breast are of the turtle type. Fig. C 1. — Feather in next to last row of lesser coverts (near outer edge of wing). C 2. — Feather from last row of lesser coverts (at about middle of row). C 3, 4. — Feathers from last row of lesser coverts (near outer scapulars). C 5. — Feather from anterior row of median coverts (first — outermost — at end of row). C 6. — Feather from mid-median row of coverts (second from outer end of row). C 7. — Feather from posterior median row or coverts (first — outermost — at end of row). PLATE 28 m •mm ■» ^■1/ ,■■■. i> 1 ■ A. Left wing of juvenal female Chamcepelia passerina. X 2. Toda del. B. Feathers corresponding to figures of the male shown in plate 27. X 2. Toda del., Sept. 1903. C. Selected feathers from a juvenal Chamcepelia passerina. X 3. Hay ashi del., 1905. ml iBaBm i THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 81 pale yellowish (text-fig. 15). A pure-gray streak of feathers is also seen along each edge of scapulars, and their feathers are not shaded with dots as are the rest, in which there is a tinge of brown. The wing-bars in this individual are of special interest, the second being very narrow and the first showing plainly its two different elements — upper and lower. In front of the oblique streak three or four rows of feathers are tinged with brown, and in front of these the small coverts are again pure gray. These are more sparely dotted in the figure. Six weeks after this bird was figured with its plumage in the state just noted, it had acquired a large portion of its second plumage, and the coverts were then of the same color as the oblique streak; i.e., the tinge of brown had disappeared and they were pure gray. This gray is a stage that was reached earlier in the oblique streak. From a young common pigeon of 17 days I have had a further illustration pre- pared of the feathers of this oblique tract as they emerge in the pin-stage (text- fig. 16). The region of these pins marks out again a region in which the feathers are nearer to the adult color. The region of the oblique tract in juvenal forms is therefore of great in- terest in studies of phenomena of reca- pitulation. The following considerations may therefore be given here: The biogenetic law certainly holds, and often in the most striking manner, in the sequence of color-patterns in indi- vidual development. But these se- quences, as a rule, appear to consist of discontinuous stages, between which, in many cases, no connecting phases are normally presented. The appearances, therefore, seem to exactly fulfil the re- quirements of the mutation theory. But the possibility of resurrecting missing links, in the manner described below, shows how illusive is the evidence from this source. The phyletic series may have been perfectly continuous, although the ontogenetic series runs off at such speed that relatively few terms appear on the surface. That a real continuity is, nevertheless, maintained is abundantly evident when, by the simple experiment of plucking a feather, and thus making room for development, we can, at will, release the stage or stages that are normally passed before nature opens the door for a new unfolding through a molt. The same continuity, moreover, is demonstrated in perfectly natural ways, or, as we might say, by nature's own experiments, since some of her regular per- formances have the value of tests most cleverly executed. In my experiments a feather is removed some time after its appearance in order that a second feather may develop which will show what advance in color or pattern has been made since the issue of the first feather. If the time of the first feather could be delayed for a few days or a week, without checking the general development of the first plumage, we should obviously have a test suited to the same need. Text-figure 16. Pin-feathers marking the region of the oblique streak in a common pigeon of 17 days. The region of the pins is the region in which the feathers most nearly ap- proach the adult color. 82 ORTHOGENETIC EVOLUTION IN PIGEONS. Explanation of Plate 29. A. Chamcepelia talpacoti. Natural size. Hayashi del., 1901. Color paler than Ch. passcrina (fig. B, this plate). Neck and breast plain, without dark centers. The primaries not perfect when drawn. (This bird was supposed to have come from Vera Cruz, Mexico. This, however, is considerably beyond the known range of talpacoti. There is a possibility of error here. The author notes that he had specimens from both Vera Cruz and Brazil. He uses only the one name — talpacoti — for all. — Ed.) B. Left wing of male Chaimrpdia pusserina. Natural size. Hayashi del., Apr. 1S97. Spots "steel blue, metallic reflections," and high coloration of coverts and primaries; the latter cinnamon-red, edged (outer) and tipped with black. C. Oblique bar in the "down" stage in male Geopelia humcraKs. Age 16 days. Natural size. Toda del., Oct. 1906. To show that the oblique band is still undeveloped and that it consists of about 19 feathers — i.e., of 13 feathers in two rows (S + 5) and then of single feathers. These feathers are only in the down stage, the feathers not having started as visible pins. The downy feathers are easily recognized in the anterior, otherwise naked, portion of the wing. I find, on mapping and measuring, two rows of down running along nearly parallel with each other and in line with the radius or anterior bone of the lower arm. These rows are separated by about 1.5 to 2 mm. The posterior row has 8 downs about 2.5 mm. apart. The entire row is IS mm. long. The second row has ."> downs, alternating with the downs of posterior row; in front of these rows are the li addi- tional downs, only 3 of which are in a row; 1 1 1« - remaining :i are anterior ami irregular in position. In older ti> see these downs it is only necessary to lift up and press to the front all the small coverts in front of the 4 rows of most regular and conspicuous coverts. The naked space occupied in part by these clowns is then seen to be more or less triangular. These downs — particularly the two rows — are destined to appear a little later, and to form "the oblique streak" in which the color-pattern makes a step nearer the adult pattern than does the pattern now situ in the coverts already out. D. Juvenal Geopelia humeralis (GH'-C 2 ). First plumage in part. Age 43 days. Shows two posterior scapulars and three inner tertials. The tertials have streaks of black on the lower edge. Explanation of Platfj 30. A. Wing of adult female Chamcepelia passerina, same bird as text-figure 4. Natural size. Hayashi del., Apr. 1903. Wing outstretched to show the extent of reddish pigment, left white in the figure on primaries and elsewhere. The wing-spots better shown than in the colored wing. Dark centers are seen in a few feathers at bend of wing. B. Adult Chamcepelia buckleyi. Toda del., after Salvadori, figure 2, to show spots of wing. The following sex comparisons are given by Salvadori: Mali . Fcmah . Upper parts vinous gray Bale brown. Under parts vinous Pale buffy brown Velvety black spots on some upper wing-coverts, scapulars, secondaries, and tertials. Spots similar. Under wing-coverts and axillaries black The same. C. Selected feathers from juvenal Chamapelia pallescens. x 2. Hayashi del., June 1907. leathers from one of six birds received from Santa Martha, United States of Colombia. This bird was in first feathers for the most part. Fn;. 3. Bosterior scapular (adult). 4. First (innermost) larger covert (juvenal). 5. Second (innermost! larger covert (juvenal). 6. Third (innermost) larger covert (juvenal). 7. Eighth (innermost) larger covert (juvenal'. (1) The first feathers have the pale edge and the pale median streak. (2) The second feathers have no pale edge or median streak. (3) A tendency for a dark cross-bar to form close to the pale edge in the larger coverts. (4) In the three inner larger coverts the spots are squarish chequers somewhat pointed distal ly -i.e., they are more like the forms seen m C. litria than the same spots in the adult. (5) The spots always remain marginal, however much reduced — e.g., the scapular. PLATE 29 B D A. Chamoepelia talpacoti. Natural size. Hayashi del., 1901. B. Left wing of male Chamcepelia passerina. Natural size. Hayashi del., Apr. 1897. C. Oblique bar in the "down" stage in male Geopelia humeralis. Age 16 days. Natural size. D. Juvenal Geopelia humeralis (GH-C 2 ). First plumage in part. Age 43 days. Todadel.,Oct. 1906. PLATE 30 iPlllPil ■ • ..... , A. Wing of adult female Chamcepelia passerina, same bird as text-fig. 4. Natural size. Hayashi del., Apr. 1903. B. Adult Chamcepelia buckleyi. Toda del., after Salvadori, fig. 2, to show spots of wing. C. Selected feathers from juvenal Chamcepelia paUescens. X 2. Hayashi del., June 1907. AKorcSCnRiBvBon THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 83 Now, nature performs just this experiment on the young of many species of pigeons, both wild and domesticated. By varying slightly the time of delay, and by operating on many different patterns, and in several regions of each pattern, nature offers a profusion of demonstrations, showing that it is not mutation, but the "law of genetic continuity" — one continuous process of development that underlies and connects the phenomena of colors and patterns in successive plumages. The intervals between stages are not periods of rest, but of active progress in transitional elabora- tion. The visible stages and the invisible flow of transitional latent phases together form a continuum, with some analogy, though only highly fanciful, to the rhizome with its visible nodal growths arranged in serial order, but appearing as isolated gene- rations, because the internodal connections lie hidden beneath the surface of the soil. One of these natural experiments is realized in this narrow "oblique tract" which, running parallel with the bones of the forearm, crosses the wing obliquely in the region of the anterior or lesser coverts. In this line, for some reason yet to be ascertained, two or three rows of coverts appear a week or two later than the other coverts of the wing, the time varying for different species. In all cases where there are considerable differences in color or pattern between the first and the second plumage, these belated feathers will appear in more or less conspicuous transitional conditions. When, for example, in the domesticated pigeons, a clear black is to succeed a blackish-red, black will predominate in the streak and thus render it quite distinct. But the black here, as a rule, will not be as strong as it is to be in the second plumage; and as the first molt leaves these feathers intact until after that plumage appears, the streak will still be recognizable, but in this case from deficiency rather than excess of black. Black pigeons sometimes produce pale or dusky red young, as a pair of black barbs has repeatedly done during the past several years. The streak in these young is a shade or two darker than other parts, and anticipates a still darker red in the second plumage. In young pigeons of the second or double-barred varieties the colors are com- monly paler than in the adult, and the gray ground-color is suffused with brown, which disappears in the second plumage. In such birds the streak shows a stronger black and a clearer gray, approximating more or less closely the adult colors. It is not in the domestic pigeons, however, that we get the more striking and instructive pictures of progressive transitional change in this region of the wing. For such views we must go to some of the wild species, in which the adult pattern departs more widely from the juvenal pattern. It is in such species, where the gap between patterns seems so wide and so abrupt as to suggest a sudden and whole- sale mutation, that we find in this tract of feathers the most complete demonstra- tions of continuity through intergraded or transitional phases. For the purpose in view we must of course have young birds accessible for con- venient observation from day to day, and such material is obtainable only from species that will breed in captivity. It is for this reason that the phenomena in question have hitherto escaped the attention of naturalists. The little ground-doves of the genus Geopelia, found principally in Australia, answer the requirements named and furnish the best demonstrations I have thus far discovered. 34 "Recapitulation of the light apical edge of G. Iranquilla and of the bars which precede the white spots or dots of G. cuneala are described and figured at the close of Chapter X. For the recapitulation of still other characters in Geopelia see the following section.- — Ed. 84 ORTHOGENETIC EVOLUTION IN PIGEONS. The Geopeliini their rei pective tages of development as lull" (I; Oeopelia tranquilla, with narrow white and black bars running entirely aero the front of the neck, with no median interruption. \'j) Oeopelia striata, with bars interrupted in the middle of the neck by ;i median treak i.i pale vinous (thi pecie tand lower than tranquilla in having barred feathei along the ide of the abdomen The two are about equal in rank). (3) Oeopelia humeralis, with "" '»"■• mi the front "I tin' neck. (I) Oeopelia cuneata, with no bars on front <>l the neck, and bars elsewhere 1 1 an im med im<> 'l"i \n l. '.', ami :; all agree in the adult in having tin' feathers of the wings and back edged by ;i narrow band of Marl., giving the whole ;i squamated or scale like appearance. No. I departs most widely (i really of generic rank), and agree with the earlier membei "I the scries only partially in the juvenile feathers. In the fii I feathers the bars inn aero the mil. in Oeopelia tranquilla, .'il o in striata, though not bo clearly marked. In 0. humeralis the bars are present but not harp, and are inlet rupted hi i In median line, <> that the juvcnal type correspond i" the adult 0. striata. Here is a curious scries of i tages plainly standing in the relation I have indicated. This group of Australian doves furni hei in omo respects the be I dei istrations of recapitulation thai I have thus far ' 1003) >li covered. Toininlncl that tho blacl ind while bare alternate i I n ind that there I no middle in tcrrupl f n il.. nock, breast, 01 ibdomon dl of which pul till bird bolow Geaptlla Iriula I • < I i i I > thai in. ni. /i | i i . i • i i ;, , | In i till would iHirn | I to tho numbor in ... ...,/, >i, /',. and Uymnojmlia Holvadnri i' 164 In Bi ■ ' thi nuuiboi ii 14, w liiah corn | I lt.li tl f ol In i >■• "i" lin I PLANAT10N 01 I'iaii 86. \ A.hili Bcnlcd dove, Scardafellu \quamoaa l od i di l . at 1 1 u Knip and Temminck, i, p. 127, pi 1 1 1 , 1808 1818 I In bird liotild Follow •'■ llghtl) more diffcrentialed on tin ving I pp I llghl i i llio tip I ii. I. i i • • dark bi I with lilaul ■ ■ il idorl, p 104.) I in l.i " I Iioav) In i "i i' i ii' • lull I . iiiinii.. I. n oris that thin bird hu 14 tiiil-fontliara, tho 10 central f equal length, tho 2 lateral Ii Idi in Irougl) 6Uigfit ! idorl i tho tail i' 1 1 in i in 12 for both pcoii of Scarilajatla, and I loulitodly rlghl i i ii I. 1 1 1 1 Igod ■■• ii i. i Im tvj itii bur of black, i oopl tl loi tail I liloli ire white The lirciinl mil Idi i pinkiidi (vinu Imilar in inea il idori phi ' g, foAthore fulntl) nn.ti led," reoall ol ■"• iiu .. 1 'i Holt'c tod feuthci from an adult Scardtifella inca Natural ize Toda del., May 1003 i i r . 1 1 ' i i 1 1 ii'iii Hi in I i. i I i. .in Inwoi I 'J I'luiii the lido, Ii "r i" nd • I ...in III I I'll. ..I i i ...... olow '" I ■ from tin uno level, closer to edgo of wing 1 1 i plain iicatl f bin dnrk and buff alternating i. I- ' ' undi i tail 1 1. I.- "i we mo tin turtli pattern red I to u thin gray i tension of tin Im Phi h il I 1 1 d " i . 1 1., ii . ...... . . i.uii ... i, i.i.i im , il tin .1 1. 1 ■ 'I|-. 1 1.. .1 .. I ■ 'I i" terminal ol ibtoriiiinnl u general!} I reduced, replaced in part by light ii ii i'ii' in I. rod replaces bluck, etc PI A I .iMMUIHm ■» ti A. Ailuli soaled dove, ScardafMa squamosa ' Toda del., after Snip and Temminok, I, p. 127, pi. ux. I' S( > S 1818. B. Ki^H. l-'.i Beleoted feathera from an adult 8oardafeUa imea Natural »"•■ Toda del., May L908. THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 89 The following detailed description of the several individuals figured will supply some of the evidence for the above conclusions: Text-figure IS. — Wing of Juvenal Geopelia striata, 18 days, x 2. Hayashi del., July 1899. This is stage corresponding with pi. 88 of Geopelia cuneata; pi. 41 of hutneralis; text-fig. 31 of tranquiUa. The third and fourth rows of coverts are so close together that they scarcely seem to be set off from the smaller coverts. At this stage I saw two pin-feathers of the oblique streak coming in, in front of the fourth row as nearly as I could tell. They came in along the same line as in G. cuneata. In this specimen the terminal yellow edging is repeated as a cross-bar in front of the black bar, and quite distinctly. In the mate of this bird the repeated yellow bar was not so plain. The black streaks are quite pronounced on lower edge of the feather, and some feathers have double streaks, upper as w-ell as lower. Text-figure 19. — Wing of Juvenal Geopelia striata, 28 days. Same bird as text-fig. 18. X 2. Hayashi del., July 1S99. Compare with 28-day wings of Geopelia tranquiUa, text-fig. 32; G. hutneralis, text-fig. 20; G. Cuneata, pi. 88. In 10 days we find the third and fourth bars have been covered by the oblique streak of feathers that had but just begun to appear in the 18-day stage, 10 unfolded feathers and 5 new pin-feathers appearing. The latter and others that will soon come into sight will unfold and lengthen until the second row is covered. We see also about 6 feathers added to the scapulars on the lower side. These are black-edged, the black being deeper than in the earlier feathers. These new feathers still show the yellow cross-bar in front of the black. They do not quite reach the adult condition. They are still only partly grown. The head, neck, and breast of two adult Geopelia striata are shown in the figures. One represents the normal type (pi. 37, fig. A), the other (pi. 38) is atavistic and more like the young (pi. 38, and text-figs. 18, 19). 90 ORTHOGENETIC EVOLUTION IN PIGEONS. The marking of the normal male shows that this species stands below Geopelia humeralis in having: (1) a smaller size; (2) a lesser area on the breast and neck that is free from bars; (3) the unbarred part pale vinous instead of gray; (4) the vinous area extends downward to about the same level as does the gray in G. humeralis; it then fades into the whitish of the abdomen {G. humeralis has lower breast only pale vinous); the vinous breast probably stands below the gray breast; (5) there are more barred feathers, and more bars to the feather. On the wings and back the bars are apical edge-bars, as in G. humeralis, and are about the same in color and appearance in the two forms. In young G. humeralis, as in this species, the black bar is not terminal, but stands within the apical pale- whitish bar. The black apical bar is therefore secondary, the pale-edge bar primary. In the adult of this species I find on the neck many feathers which retain the white apical edge, while some others show the apical black bar in all stages of obsolescence. The pale or white edge of the neck-feathers is therefore a still later modification, the feathers having lost the juvenal pale edge and are now in process of losing the black edge, which is homologous with the black edge of the wing-feathers. The neck- feathers are, then, higher in development than the wing and back feathers. As we pass up from the back to the neck we see the black edge is at first present in full width; on the lower hind-neck we find it looking a little thinner and lower; Explanation of Plate 37. A. Adult male Geopelia striata, x 1.4. Hayashi del., Nov. 1902. Shows the extent of the unbarred breast region in the normal type of this species. Compare with the more complete barring of juvenal and adult atavistic female (pi. 42). (See text.) The feathers are all shown with white edges, as they appeared to Mr. Hayashi, but the black apical edge is for the most part present, though inconspicuous. 1 to 7. Separate feathers from normal G. striata. A. Upper part of mid-back of neck. Bars are only 3, the basal one very shadowy. The mid-back neck-feathers are all less differentiated; they have fewer bars and the color of these is intermediate between that of body and wing feathers and of the feathers of the side of the neck, where differentiation is sharpest. The black bars are here only dark brown and the white bars are pale buff. B. Shows the height of differentiation in mid-back of neck. C. An ordinary feather from middle of back of body. It has black apical edge-bar, and just inside this an obscure buff bar. D. Middle height of neck, right side, front edge of barred feathers. Front half of feather is not barred, except that the first black bar runs over the mid-rib a little. This half is vinous. There are four half-bars. The apical black bar is only barely recognizable to the naked eye (on the barred side), and extends just a little over the mid-rib and soon vanishes. The artist at fust failed to sec and draw the black apical bar. E and F. Middle right side of neck, where bars are best developed in number and color, the pale bars being almost white and the dark bars black. Figure E has a thin black edge, but figure F has only a trace of it on the left side. G. Just below middle left side of neck, similar to figure F. To the naked eye there is no apical black, but with a lens I can detect just a touch of blackish on the tips of some of the barbs. Note that this species has the whole side of the body and under wing-coverts multibarred. Explanation of Plate 38. A. Juvenal Geopelia striata, x 1.5. Hayashi del., Oct. 1902. This first plumage shows the bars completely circling the front of the neck, but they are not so well marked in the middle region, which is dl stined to be even unbarred and of vinous color in the second plumage. This condition of bars is permanent in one species, T— !Z; -t-^ ~ o O id r. • p-t * ,a a QQ T 03 3 QQ 03 03 •v. 03 • w iC x •s e %> ^ .*"*■■ ■ v m^, M "■%. M &&$ isa Mm " ■";» ' A 1 * ^S ' A. Adult male Geopdia humerdis. X 1.4. Hayashi del., Oct. 1902. B. From near front edge of neck in "ring" region (left side). C. Juvenal Geopelia humerdis, age 28 days. X 1.4. Hayashi del., Nov. 1902. D. From side of neck (of above bird) near bend of wing. X 2. E. From middle of breast. X 2. ■ PLATE 41 ;ii?%i i i ., a - -^ ■ ■ w m "?% - Bfc& •: B $ E A. Wing of Juvenal Geopelia hvmeralis, age 11 days. X 1.5. Hayashi del., Aug. 1899. B. Wmg of Juvenal G. humeralis, age 18 days. Natural size. Hayashi del., Aug. 1899. C. Four inner tertials of G. humeralis. Natural size. Hayashi del., .July 1907. D. Three inner tertials of Zenmdmra eardmeiuiB (PP1). Natural size. Havashi del., July 1907 E. Three inner tertials of Z. carolinerms (FP2) . Natural size. Hayashi del., July 1907. WfairrtCoBm.Bire THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 95 Spots or marginal streaks are recapitulated in the tertials, scapulars, and coverts of the juvenal wings (pi. 41, A and B, and text-figs. 20, 21). A comparison of these streaks in the tertials of humeralis and the mourning-dove (Zenaidura) is made possible by figures 3 to 5 of plate 41. The homology of the wing-chequers of the juvenal geopelias with the chequers in Columba livia has been pointed out in the preceding chapter. In all of the geo- pelias 41 these chequers are completely lost with the first feathers. 42 In the ontogeny of the geopelias we see the different species all taking departure from a common color-pattern, consisting of longitudinal spots or chequers, dis- Text-figure 21. — Wing of juvenal Geopelia humeralis, age 10 weeks. Natural size. Havashi del., Aug. 1899. The first row of coverts retains only 4 feathers seen in the previous (30-day) stage, and these are being covered by new black-tipped feathers. Only 3 old feathers remain in the scapulars. Text-figure 22. — Wing of adult Geopelia humeralis. Natural size. Hayashi del., June 1S99. Drawn for comparison with wings of the young. tributed uniformly to all the feathers of the wing. We see this common and earlier pattern retained for only a few weeks, the later pattern developing in two widely diverging directions and leading to two types of generic rank. The Phabin.e. The relation of the color-pattern of two of the genera of this subfamily has been already described — Ocyphaps in Chapters II and III (pis. 8, 14, 15, 16; Phaps in Chapter II (pi. 6). (A short treatment of three other genera — Calopelia, (Ena, Tympanistria — of this group is extracted from a short paper without illustrations published 43 in 1907.— Ed.) 41 Geopelia or Stietopelia etineata is described and figured in Chapter IX. — Ed. 42 It has been observed that the feathers of the oblique streak do not make their appearance at exactly the same age in the several species of Geopelia. The order is as follows: (1) G. tranquilla: streak out at 17 days and a second row with black tips begins; (2) G. striata: streak just begins at 18 days; (3) G. euneata: two rows well started at 19 days; (4) G. humeralis: begins about 19 to 21 days; at 28 days one row and three more black-edged feathers (some with very narrow light edge). 43 Avicultural Magazine, London (Journal of the Avicultural Society), n.s., vol. 5, No. 6, April 1907. — Ed. 96 ORTHOGENETIC EVOLUTION IN PIGEONS. In a review of my paper, "The Problem of the Origin of Species," Dr. A. G. Butler comments 44 as follows on my conclusion that the bars in pigeons have been evolved from chequers: This is certainly not the conclusion to which a study of the nestling plumage of some at least of the African doves leads me, inasmuch as the conspicuous spots on the scapulars and inner secondaries appear in the adults with the disappearance of the juvenile bars. I presume that the maiden-dove (Calopelia puetta) , recently described by Dr. Butler, 45 was one of the "African doves" he had in mind. This species is pictured as having only three "metallic spots on the inner greater wing-coverts and largest scapulars." An immature bird described by Captain Shelley in 1883 (Ibis, p. 322) is referred to as having "black bars on the scapulars, wing-coverts, and secondaries." These "black bars" of the young are not further described, and I am therefore in doubt as to whether they represent rows of cheqxiers or the higher stage of develop- ment seen in apical crescents, a form characteristic of the inca-dove and the geopelias. The mature color-pattern in Calopelia does not differ widely in essentials from those seen in some American species, e.g., the mourning-dove (Zenaidura caro- linensis) and the zenaida-dove (Zenaida amabilis) . In the young of these doves we find many typical chequers, more or less evenly distributed over the whole wing; in the adult we have only a few of these spots left, and left in the same region in which the three spots of the maiden-dove are located. The obliteration of the spots in the American birds has, however, not yet been carried quite so far as in the African dove. In the mature mourning-dove we find not only a larger number of visible spots, but also many concealed vestigial spots. Zenaida has carried the reduction of spots somewhat farther, and stands only a little behind the maiden- dove. Our white-winged pigeon {Mclopelia leucoptera) has practically completed the deletion of spots, only a very few vestigial traces being discoverable in a single specimen obtained from Jamaica. I do not find such vestiges in white-wings from Mexico and Arizona. In a male cape-dove (CEna capensis) that has just come to hand I find indica- tions of a still closer correspondence to the Zenaida and Zenaidura types. In this dove (pi. 42, fig. A) there arc two very short bars, one on the tertials (with one spot on the right wing and two on the left), and another on the inner long coverts (with three spots on the right wing and four on the left). These black ("steel-blue") spots are subtcrminal squarish blocks on the outer webs. On the inner webs of the tertials bearing the bar-spots I find elongate black spots reduced to narrow marginal streaks in most cases. A tertial with such a streak on the inner web extending nearly to its tip and a bar-spot on its outer web at a considerable distance from the tip presents a picture quite characteristic of the zcnaidas and the mourning- doves. So close and peculiar a parallel in the mature patterns of these doves would lead us to expect fully as close a resemblance in the juvenal patterns; but Salvadori's description of the young cape-dove does not seem to confirm this anticipation. "The wing-coverts," as he reports, "are grayish-brown with blackish bands and whitish-buff apical spots." 44 Avicultural Magazine, Dec. 1906, p. 71. « Ibid., June 1900, p. 251 THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 97 According to Dr. Butler (Avic. Mag., n. s., Vol. II, p. 101), the young tambourine- dove ( Tympanistria tym pan i stria) is similarly marked : "All the feathers of the wing and tail are of a bright coffee-brown color with broad subterminal irregular trans- verse black bands." The figure given in a later volume of the same magazine (Vol. IV, page 308) makes it clear that the young tambourine rises to a stage of irregular cross-bars analogous to what is seen in the young inca-doves (Scardafella inca) and geopelias, in which we find the feathers edged apically with a pale straw-color (very narrow in inca but conspicuous in geopelias), followed within by a blackish cres- centic bar, and then one or more quite broken pale dull "bars" (too irregular to be described as bars — mere suggestions of bars). In the mature state the pale apical bar is lost and the blackish crescentic bar becomes terminal. This form of barring, so far as the black crescent is concerned, is something later in evolution — as I have elsewhere (in the present chapter) made clear — than the lateral chequers of the mourning-dove, passenger-pigeon, etc. In young geopelias (e.g., pis. 38 and 41 and text-figs. 18 to 21) we have transient lateral chequers in the tertials and longer wing-coverts, and in such continuity with the black crescents that the latter must be regarded as derived from the former. Even in domestic pigeons we frequently see chequers reduced to black crescents. In the young inca-doves (pi. 34, figs. C and D) these same transient chequers are recapitulated on the tertials and the long coverts. Although not so black as in the geopelias, they are yet plain and unmistakable homologues. Only two or three of the inner long coverts have this vanishing chequer as a long lateral streak on both the inner and outer edge of the feather. On the remaining feathers of the row the mark appears only on the outer web, and becomes weaker and narrower as we de- scend the row, until on the outer two feathers it is wholly lost. The recapitulation of the marks in the inca and its South American allies, and again in all the geope- lias of Australia — even in the diamond-dove, standing at the extreme upper limit of evolution thus far reached in this interesting genus — gives us a very important link in the sequence of phyletic stages. I venture to predict that the young tambourine, the young cape-dove, and some of their nearer allies will be found to have more or less plain traces of the transient marginal streaks seen in the tertials and long coverts of the inca-dove, and perhaps also dull spots on the outer webs anticipating the spots of the adult. Without going into the evidence here, I may say that I have fully satisfied myself that the lateral spots or chequers are derived from the turtle-dove spots, such as are still seen in Turtur orientalis, and, in a somewhat reduced form, in the European turtle-dove (T. turtur). The mode of derivation was by splitting the original central spot into halves. The splitting began at the apex of the feather, a short, wedge-shaped area of lighter color (i.e., reduced in pigment) appearing at this point and extending more and more inward along the shaft, until the divided halves became two separate spots or chequers, more or less pointed at the distal end. The feather thus became double-spotted. Typical "wedge-shaped areas" are not rare in domestic pigeons with the chequered pattern, and they are very characteristic marks in the wing of the guinea-pigeon {Columba guinea), where they are described as "triangular white pots." They are seen again as a specific character in the spotted pigeon (C. maculosa) of South America. They occur also in the scapulars of C. albipennis of Peru and Bolivia. 98 ORTHOGENETIC EVOLUTION IN PIGEONS. This simple mode of converting the turtle-dove spot into a pair of lateral chequers is carried out most perfectly in the posterior scapulars and in the inner secondaries and large coverts of the wing of the domestic pigeon. As we pass down- ward towards the lower (outer) edge of the wing, the chequer on the inner web diminishes in size more or less rapidly and may be continued to or near the outer edge of the wing. Now, chequers are the elements out of which wing-bars arise. "Rows of chequers," as we see them in the young mourning-dove or in the young passenger- pigeon, or in the chequered rock-pigeon (Columba a finis), may of course be called Explanation of Plate 42. A. Cape-pigeon, female Oena capensis. Toda del., after Temminck and Knip, Les Pigeons, i, p. 117, pis. liii and liv, 1808-1813. x ? Temniinek and Knip picture both male anil female. A very interesting species for comparison with the crescent-barred geopelias. It is much more highly differ- entiated than any of the geopelias in that: (1) The male is differently colored from the female, having its (a) forehead, throat, front neck, and breast black; (/)) the occiput and hind neck are pale brown shading into pearl-gray on the sides of the neck (the female has the whole head and neck and breast pale brown, becoming lighter in front half of head and throat) ; (c) "the inner upper wing- coverts have large steel-blue patches" (Salvadori), which are not seen on the female. These spots remind of the Florida ground-dove. (2) The dark crescents have all disappeared except in the upper tail-coverts (possibly the few black bands on the lower back are derived from the crescents), where they are still preserved in both sexes. (3) The great length of the middle tail-feathers. (4) The 2 (4 in female? Temminck says 3 bands) black bands across the lower back. According to Salvadori (p. 503), the young has dark crescent-bars on crown and neck, and the wing-coverts are grayish-brown, with blackish crescentic bars and light buff edges. The pictures (of Madame Knip) of the adult male and female show that the crescents arc apical in the male and subapical in the female, the female seeming to keep the pale edge of the young. B. Right wing of adult female tambourine-dove, Tympanistria tympanistria. Natural size. Drawn from U. S. Nat. Mus. No. 116316. Here we find in the tertials 2 plain spots and 1 very weak spot, ordinarily concealed, and perhaps not always present. In the long coverts 3 plain spots — one more than in the Juvenal male, an indication that female has more spots than male. In the first row of median coverts we have two concealed spots, seen by lifting the overlying row. These were not found in the Juvenal male. (The single Juvenal bird examined is elsewhere stated (see figs.( ' and D) to be a female. I am quite unable to understand this and the earlier reference to a "Juvenal male"; unless perhaps a figure or descrip- tion — by Dr. Butler (?) — of such a bird was at hand.— Ed.) Again an indication that spots are more numerous in the female than in the male. No spots in the scapulars, and no spots or streaks on the inner webs. These spots extend towards the base of the feather, as in the Juvenal. There is probably more pigment in the adult (permanent) spots, but I should expect to find in the young female a little larger number of spots than in the adult. C. Three inner tertials of Juvenal female tambourine-dove. Drawn from U. S. Nat. Mus. No. 117807, July 1907. Shows rufous apical marks (of 1 to 1.3 mm.) followed by a dark bar, darkest next to the apical mark. The dark bar is narrow (1 mm.) at the margins and wider (up to 3 mm.) near the feather's shaft. This dark bar is followed by a narrow irregular rufous bar. Shows also black chequer on the outer web; these are squarish, on the margin and about 10 mm. from tip in first feather, 15 mm. in second feather. Only two chequers in this row; size is about 7 by mm. D. Two inner long coverts from same bird as above. In the long coverts I also find only 2 spots, but these extend farther towards the base of the feather and thus show thai the chequers are derived from the dark center of the turtle-dove feather. For a description of general ground-color see Butler, Avic. Mag., Aug. 1000. E. Wing of adult male Chalcopelia afro. Natural size. Hayashi del., Aug. 1901. This species forms one of the color-types connecting the turtle-doves with Geopelia. ( 'ompare with Peristera dnerea (pi. 31) and with open wing of another adult male ('. afra (pi. 35. fig B), where the legend applies also in part to the present figure. PLATE 42 A. Cape-pigeon, female Oena capemsin. Toda del., after Temminck and Knip, Les Pigeons, I, p. 117, pis. Lin and liv, 1808-1813. B. Right wing of adult female tambourine-dove, Tympanislria tympanistria. Natural size. Drawn from U. S. Nat. Mus. No. 116316. C. Three inner tertials of juvenal female tambourine-dove. Natural size. Drawn from U. S. Nat. Mus. No. 117867, July 1907. D. Two inner long coverts from same bird as above. Natural size. E. Wing of adult male Chalcopelia afra. Natural size. Hayashi del., Aug. 1904. .:-:».*«?* THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 99 "bars"; but in that case we have to remember that they are yet a long way from the finished bars of such birds as C. livid, the bleeding-heart (Phlogoenas) , many Peristera?, and nearly all pigeons with bars in the adult plumage. So long as the whole wing remains more or less uniformly marked with distinct lateral spots, con- fusion may be avoided by describing it as spotted or chequered. C. affinis has been so described, and Ectopistes, which has rows of spots both in young and adult stages, is never described as a barred pigeon. American and European ornithologists have generally, I believe, adhered to this mode of description. Salvadori, for example, in his Catalogue of Pigeons, speaks of Ectopistes as having "scapulars, tertials, and median wing-coverts with velvety-black spots," and of Chamcepelia as "marked with blotches of a steel-blue with violet reflections." In all typical turtle-doves we have rows of spots, but we never think of describing them as bars. The distinc- tion between spots and bars has become so fixed in the breeder's terminology that I did not realize the danger of any confusion or misunderstanding on this point before reading Dr. Butler's very courteous review. Had it been possible to illus- trate my paper with plates my meaning would probably have been clear in regard to the evolution of wing-bars from chequers, at least in so far as concerns the rock- pigeons and their descendants. The typical wing-bar of the adult bird, as I conceived it, represents a specific regional mark — a continuous band of color on a uniform ground of contrasting color. In such a bar the individualities of the elements are submerged in the individuality of the bar. When, therefore, as in Columba affinis, we meet with rows of chequers, and find that the posterior two rows are the homologues of the two black bars of C. livia, and, further, that this two-barred condition is reached in domestic birds through the obliteration of the anterior rows of chequers and by cutting down the chequers in the remaining two rows to outlines that in each row flow together in a single band, we are warranted in saying that bars are evolved from chequers. If we speak of the original rows of chequers as bars, then it becomes necessary to distinguish between juvenal or primary bars and adult or secondary bars. The direction of derivation remains the same. If we find cases in which the deletion of spots has been carried still farther, until only a few of the larger spots in the poste- rior rows remain, as in Zenaida and Zenaidura, we may then say that "spots" are the true homologues of typical bar-elements in other species. I believe that the few spots in the wing of the adult Calopelia are such homo- logues, and that they are preceded in the juvenal plumage by corresponding marks which have been overlooked. I am aware that in the young of this species the scapulars, wing-coverts, and tertials are said to be "barred with black." These juvenal bars are probably of the same nature as those seen in the young tambourine, the relation of which to chequers has already been indicated. These remarks will make it sufficiently clear that my view does not contradict Dr. Butler's observations on African doves. In the above remarks 46 1 ventured some suggestions as to the phyletic sequence of color-marks in the tambourine-dove and its nearest allies. These remarks on this dove were based on the following description by Dr. A. G. Butler, 47 in which 18 The following statements were written several months after the preceding account was published. In the mean- time, the author had been able to see and study a few specimens of the tambourine-dove. — Ed. 47 Agricultural Magazine. Aug. 1900, p. 309. 100 ORTHOGENETIC EVOLUTION IN PIGEONS. was included the statement of Mr. Oberholser. Speaking of the young, 19 days old, Mr. Butler says: "Colouring of upper parts largely bronze-brown, but the feathers barred with buff and black; flights reddish-chestnut; tail feathers vinous brown, the outermost feathers white. Forehead and broad eye-brow streak buff; the feathers at sides of crown standing in curved rows so as to produce a sort of divided crest, buff -brownish ; ear-coverts and cheeks leaden gray; the former apparently narrowly barred white and buff, but this appearance is probably due to the sheaths still remaining on the feathers at this part of the head; sides of neck, throat, and breast huffish brown with narrow blackish bars; abdomen white; bill dull black; feet dark leaden gray with a faint sub-tint of flesh color; the eye was too sunken to describe. "In his account of 'Birds Collected by W. L. Abbot in the Kilmanjaro Region, East Africa' (Proc. U. S. Nat. Mus., Vol. XXVIII, p. 843), Prof. H. C. Oberholser describes the immature tambourine-clove compared with the adult as follows: 'Upper surface of the body more rufescent; forehead grayish, slightly tinged with tawny; crown washed, the back and rump barred with rusty; wing-coverts and secondaries duller, as well as rather paler, with mottlings and some bars of dark brown and tawny, the secondaries with a dark subterminal bar, sides of the head shaded with ashy and brownish; anterior lower parts more or less barred with dark brown and ochraceous; lower tail-coverts with tips, ami simic- times bars, of tawny. This would probably be an intermediate stage between the nestling and adult plumage!' " Explanation of Plate 43. Figs 1, 2, 3. Three wing-feathers of adult female green-winged pigeon, Calcophaps indica. x 2 Toda del., Nov. 1903. These feathers, plucked in May 190:?, show transverse bars. The green portions of the feathers are left white, except for the bars. It looks as if the turtle pattern was the foundation, and as if the light edge had spread inward, becoming green (sometimes bronzy red) ; or one might think that the exposed part of the black had changed to green. The change, whatever its nature, begins at or near tip and progresses towards base. The upper feather is an inner long covert (left side). The middle feather is an inner long covert (the next below). The lower feather is an upper (inner, left) tertial. The bars cross both webs, and are on green and dark (blackish-gray) color as well. At the tip they more nearly form a right angle with the shaft; inward they meet at a very obtuse angle. They are visible on the distal two-thirds of the feather. Figures 1 to 5 all show the "fundamental bars" to which reference in the text has several times been made.— Ed. Fig. 4. A posterior scapular of a male Phaps chalcoptera. Natural size. The "barring" is very plain, but stands out more conspicuously in the drawing than in the feather. The bars are pretty regular in their distance apart— about 14 of them on the distal two-thirds of feather. They are symmetrical and form an obtuse angle at the shaft. In the long coverts they are also seen, though not so plainly. The spot or chequer is quite strong in the scapulars of this particular male; in another male the scapulars were without spots. Fig. 5. From a femule nicobar pigeon, Calcenas nicobarica. x 1.4. One of the long feathers from the back of the neck, showing regularly spaced cross-bars. The feathers of the back were more faintly barred. Figs. 6, 7, 8. Three wing-feathers of a juvenal (7 weeks) Leptoplila sp.? x 2. Hayashi del., Oct. 1903. (The exact location of the feathers drawn is not indicated. — En.) The coverts of the wing show rufous apical marks (and in some this is followed by a subterminal bar), and this is soon followed by an intimation of a rufous bar and a midrufous streak. The subterminal bar was later found to be present on the tertials of another Leptoplila (reirhmhwhi), and here the apical mark widens at the middle of the tip; on the innermost tertial this apical white is prolonged to a distance of 4 to 5 mm. as a shaft-streak. PLATE 43 / m i My 6 '■■: Figs. 1, 2, 3. Three wing-feathers of adult female green-winged pigeon, Chalcophaps indica. X 2 Todadel., Nov. 1903. Fig. 4. A posterior scapular of a male Phaps ehakoptera. Natural size. Fig. 5. From a female nicobar-pigeon , Cakenas nicobarica. X 1.4. Figs. 6, 7, 8. Three wing feathers of a juvenal (7 weeks) LeptoptUa sp.? X 2. Hayashi del Oct. 1903. THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 101 Although I had not yet seen this species, I felt safe in assuming that the develop- ment of its color-pattern would follow the same sequence that I had found to be general among pigeons. Accordingly I ventured to predict that: "The young tambourine, the young cape-dove, and some of their nearer allies, will be found to have more or less plain traces of the transient marginal streaks seen in the tertials and long coverts of the inca-dove, and perhaps also dull spots on the outer icebs anticipating the spots of the adult." Through the kindness of Mr. Rathbun, I have received from the U. S. National Museum three specimens of this dove — a pair of adults and the "immature" speci- men described by Mr. Oberholser. The latter, as a moment's examination shows, not only verifies my prediction, but also reveals some other features that still further justify the comparison I made with the zenaidas, incas, and geopelias. In both the tertials and long coverts (pi. 42, figs. C and D) I find a small number of black chequers on some of the outer webs. These chequers extend very consider- ably farther toward the base of the feather in the coverts than in the tertials and thus show that the chequers are derived from the dark center of the turtle-dove pattern. The light (rufous) apical edge, the subterminal dark bar, followed by another rufous transverse bar, all reminding of the geopelias, are all found in these juvenal feathers. It is probable, too, that the spots are more numerous in the adult female (pi. 42, fig. B) than in the male. The wing-bars of Chalcopelia afra are figured in plates 35 (fig. B) and 42 (fig. E). Shadowy spots or chequers are found on the tertials and longer coverts. These remnants of spots are mostly on the outer webs. The cross-stripes on the back, or rump, of these birds are formed of black-tipped feathers — the homologues of the black crescents. The iridescent regions of three wing-feathers of Chalcophaps inclica are indicated in plate 43 (figs. 1 to 3). The young of this species seem to have the turtle-dove pattern. Mr. D. Seth- Smith 48 says: "On leaving the nest the young birds are dark brown, the feathers of the head, breast, and wing-coverts being broadly margined with chestnut. Some of the lesser wing-coverts are green, and there is a distinct greenish tinge to some of the secondaries and their coverts." In Phaps chalcoptera (the bronze-wing) we find bars developed over nearly the whole wing; 45 in Phaps elegans (the brush bronze-wing) there are only two posterior bars. In Geophaps smithi (Smith's ground-dove) we find only a few feathers of the two posterior bars retaining a bronze color (see pi. 48, fig. C). In Ocyphaps (crested pigeon) we have many bars, strongly reduced, and in Lophophaps we get a further reduction. Lophophaps leucogaster (text-fig. 9) stands above Ocyphaps also in having no elongated marks or marginal streaks on the long coverts. (Reference to this pat- tern is made in Chapter II. — Ed.) ,8 Avicultural Magazine, July 1900, p. 277. "A scapular showing a marginal chequer and the "fundamental barring" is shown in pi. 43, fig. 4. 102 ORTHOGENETIC EVOLUTION IN PIGEONS. The Geotrygonin^e. The Geotrygoninae form a large subfamily of pigeons (see table 2). These are all birds that run on the ground most of the time, and their behavior in fighting, in cooing, and in hiding with tail raised and head down stamps them all, so far as I have seen them, as of common descent. It is in this group that we find the Leptop- tila, Osculatia, Geotrygon, and Stamcenas of North and South America; the Phlo- goenas of the Philippines, the Leucosarcia of Australia, the Haplopelia of the Ethio- pian region, the Eutrygon and (Jtidiphaps of New Guinea. The group then has a wide distribution, and furnishes another example of birds that have departed from the turtle-dove pattern, mainly by loss of spots. In a Geotrygon (sp. ?) from Brazil I find dark centers reduced to a minimum in the under tail-coverts and on the under surface of the body generally. The long coverts of the wing have dull dark centers and some of them show a light (rufous) edge. In young geotrygons I have seen a dull reddish-brown mid-streak in the upper wing-coverts. This streak is of the same color as that of the same streak found in Melopelia. Leptoptila shows characteristics — in courting behavior and in voice — that seem to match well with those of Calopelia. Dr. Butler 60 makes note of the agreement in behavior. The juvenal upper wing-coverts (pi. 43, fig. 6) of Leptoptila are of interest in showing the light mesial streak 61 which, as we have already noted, is the means of ^dividing the dark turtle-dove center into two lateral chequers. Phlogoenas luzonica — the bleeding-heart pigeon — has three plain wing-bars in the posterior part of the wing. A fourth bar is so covered and irregular as not to appear, but two 'middle coverts show that the bar-elements are here continued, though reduced in sharpness. The wing-pattern thus strongly reminds of conditions in Columba livia. Leucosarcia picata, the white-faced pigeon, has the turtle-dove pattern in the adult feathers; such feathers are, however, limited to the larger portion of the under surface of the body. A series of 19 feathers, taken from along the right side of the body, from the breast to the under tail-coverts, are shown in plate 44. The pattern exhibited in this series of feathers deserves further consideration. In the flank feathers (12-14) and under-tail coverts (15) we find the typical turtle pattern. The same pattern, but less perfect, is shown in the under wing- coverts (16-19). Passing forward on the abdomen, we see this pattern so gradually modified, by reduction of the dark center, that we feel sure, when we reach the "Avicultural Magazine, June 1906. 51 This median streak is of the same nature as the apical spots of the Zenaida mutant ! (See Chapter IX. — Ed.) Explanation of Plate 44. Selected feathers from white-faced pigeon, Leucosarcia picata. x 0.8. Hayashi del., May 1903. 19 feathers along the right side, from breast to under tail-coverts; 16 to 19 are under wing-coverts. No. 1. From lower breast, right side. No. 15. An under tail-covert; illustrates all of the No. 2. From close behind No. 1. „ „ tu "; tIe Pattern. „ „ ,. _ ,. ,. . , . . , .. Nos. 16 to 19. under wing-coverts. These are in 4 No. 3 to 11. Continue the series and reach about half- transverse rows way to the leg, or to about the middle of the No ]6 From middle of first (anterior) row. abdomen. No. 17. From middle of second row. Nos. 12 to 14. From the flank feathers lying across the No. 18. From middle of third row. leg and reaching to under tail-coverts. No. 19. From middle of fourth row The color of 1 to 15 is brown-gray for nearly two-thirds of the feather's length. The distal third is white. Feathers 10 to 14 have a faint tinge of buff in the white. PLATE 44 1 m Ik JL I 7 i ;■' ■1 ;, ' 12 11 - 1 i ** 13 14 10 , 1 t'fflji I ' 'juip w *> , -^m '0? '"-"'■'? • 18 T 15 Selected feathers from white-faced pigeon, Leucosarcia picata. X 0.8. Hayashi del., May 1903. AJttnS THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 103 feathers with roundish dark spots in a white ground, that these spots are remnants of the original dark center. The most anterior feather (No. 1) shows a mere trace of a spot. But the whole center is not lost — it is only the distal third that suffers reduction. If we begin with 1 and pass backward through the whole series to 15, we could say that the gray of the basal part of the feather is gradually increased, and that the steps in evolution led from a dark feather to one with a dark base and a white end; then to the establishment of a dark spot in the white end-segment, and by increase of dark pigment to a coalescence of the basal pigment with the spot, and finally to the turtle pattern in 15. We should then incline to think that the next step would be to complete the series by extending the dark to the very edge of the feather, and thus reach the highest stage of a uniform color. This mode of reasoning has fre- quently misled investigators. 52 According to my view, the original pattern was the turtle-dove pattern. On the upper parts the light edge of the feathers has disappeared; on the lower parts the original pattern is best preserved posteriorly, and more and more modified as we go forward, so that 1 would represent a higher stage than any of the stages that follow. That this is the correct view can be shown by a general comparative survey, by comparing the juvenal with the adult patterns, and by studying the history of transverse bars in such forms as the flicker (Coloptes auralus, pi. 59). Salvadori (page 008) says of the young of this species: " It is similar to the adult, only browner on the wing, and without the dark centers to the feathers of the sides." By "dark centers" Salvadori means such spots as are seen in figures 3 to 8. I find the feathers along the flanks spotted, but the spots are weak and not at all con- spicuous. The middle region of the abdomen and lower surface is whitish without spots, showing a faint trace of buff. The feathers of the side, just behind the legs, have a stronger tinge of buff. The color of the young is closely like that of the old bird, but has a plainly deeper tinge of brownish. The feathers of the forehead are a whitish buff, which becomes lighter from the beak over the eye. The under tail-coverts are nearly buff color, and are without spots. I find the light apical edge. It is, however, very weak and thin — scarcely noticer able even at close view. It is present in some three or four rows of wing-coverts on the feathers which lie outward (downward) from the middle of each row; above this level (on the inner feathers) it is not visible in these rows, and on the lesser coverts I miss it altogether on all the feathers. The secondaries and primaries are pale light buff along the outer edge and around the tips; this pale edge grows less in passing upon the inner webs. Three or four outer tertials have a trace of the apical mark from 0.25 mm. to 0.5 mm. in width. In the long coverts, where these marks are somewhat plainer than in the median coverts, the greatest width of the mark is about 0.5 mm."'" "E.g., Hacker, Keeler, and others. (See Chapter VII. — Ed.) 63 The general color-pattern of the Caleonadins — the last subfamily of the Peristerida? — has not been described by the author. He has noted however, that "the while tail of Catenas nicobarica is a character found only in the adult; in the juvenal stage — called Cat. gmddia: by Gray — the tail is greenish-black (see A. G. Butler: Knowledge Revealing Ignorance, Avic. Mag., Nov. 1910, page 46)." Some will be interested to know that the sentence just quoted seems to be the last statement written by Professor Whitman into his manuscripts. The modified neck- feathers of this species (Cat. nicobarica) are figured in plate 33, Vol. II, and a single feather is herewith reproduced in plate 43, fig. 5. Note. — The manuscripts or records used in the preparation of this chapter were found in folders designatedas: follows: A 4, A 0, A 16. Em 1, Em 4, H 2, K 12, Misc. 6, 04; 005, 7, 9, 11, 16, 17, 18, 20; R 1, 12, 13, 16; Sh 24 3 /l W 4, 5, 10, 11, 13; WW 6, XW 2; Z 3, 4, 10, 11.— Ed. 8 CHAPTER VI. THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS (CONTINUED). THE TURTLE-DOVE PATTERN IN THE TRERONIDit (See table 1, page 66.) The TreroninjE. In Sphenocercus sphenurus, according to Salvadori (pi. 5) and Bonaparte, the long coverts are green with yellow edges; the secondaries are slate-black, edged with yellow; the tertials are green without bright edges. In many species the primaries are edged with yellow or a bright color. In Siebold's green pigeon (Sphe- nocercus sieboldi) of Japan 1 both male and female have the under tail-coverts dark (olive-green) centered and edged with yellowish white. The dark-centered feathers extend forward to the legs. The long coverts of the wing also have blackish centers and yellowish edges. The PTILOPODIN.E. In Ptilopus dupelit-thourarsi the upper parts of the adult bird are mostly green, but the larger scapulars and the tertials have dark (deep blue) triangular centers, which are pointed behind. The shape of these dark areas is that of the dark centers in the European turtle-dove. Five of these spots are shown in a figure by Bona- parte. 2 The secondaries and greater wing-coverts are bright green, edged with yellow. Yellow here takes place of the reddish of the turtle-dove and green the place of dark centers. The scapulars and tertials have centers of deep blue and are edged with golden green. Unmistakably the turtle-dove pattern is preserved here in a case where the colors are most ornamental. Many of the tropical green- winged pigeons of the Treronidae have the tertials and one or two or three of the rows of longer coverts strongly edged with yellow, the central part of the feather being dark. Ptilopus xanthogaster, according to Salvadori, 3 has green back, wings, and tail. The gray feathers of the upper breast are bifid — a condition already noted in the neck-feathers of both Peristeridae (Spilopelia) and Columbidse (Columba guinea). The feathers of the thighs and flanks have green centers and yellowish edges. The smaller and median wing-coverts have green centers and golden-green edges; the greater coverts and the secondaries have yellow edges. The scapulars and inner secondaries have deep-blue centers, edged with yellow. These deep-blue centers are not pointed, but fill out the feather as does the black in the Japanese turtle-dove. Ptilopus chrysogaster i has all of the wing-feathers edged with yellow where the turtle-dove has edges of buff color. The feathers of the crop, or upper breast, are bifid. Ptilopus marice is one of the most highly colored and variegated of pigeons. In this species the male differs widely from the female and the young. In the adult male we see nothing that points decidedly to the turtle-dove mark, but Salvadori 1 I have a mounted specimen brought from Japan by Mr. Hayashi. 2 Iconographie des pigeons, pi. xvn. 3 This is ThouaTsitreron diademaia of Bonaparte (Iconographie, pi. xvm). 4 The descriptions of Ptilopus are based on the statements and figures of Salvadori and of Bonaparte; the latter has figured P chrysogasUr in pi. xxix; P. mariir in pi. xvi. 104 PLATE 45 Right wing of adult crowned pigeon, Goura coronata. X 0.6. Hayashi del., Feb. 1904. From one of two birds obtained from dealer in autumn, 1900. General color bluish slate-gray, wings darker. Middle and bases of longer wing-coverts (which fall in the center of the plate) are white, very broadly edged with chestnut. The anterior coverts show dark-slate, pointed centers, with broad chestnut-colored tips. AH jpS [ q£; in mart THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OP PIGEONS. 105 (page 89) notes that "this species passes through many stages before assuming the full dress." The young is so different from the male that some have found it diffi- cult to believe it belongs to the same species. The young, however, does resemble the adult female. The wing is green, but the smaller coverts are edged with a lighter bronze (?), presenting an appearance that closely resembles, in general effect, the wing-pattern in the turtle-dove. The dark (green) centers are more or less pointed, especially on the larger scapulars and inner tertials. The long coverts have green centers and are edged with yellow, and the same is true of the secondaries. The breast-feathers have green centers and light-grayish edges or tips; the upper breast- feathers are bifid and bear whitish tips; the feathers of the lower breast have green centers and wide grayish-yellow edges, just the turtle-dove type of pattern. 5 THE TURTLE DOVE PATTERN IN THE GOURIDj