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. 



