THE PROBLEM OF ORGANIC DEVELOPMENT. 183 



undoubtedly lost it. 2 It differs in different species in width, in color and shade. 

 Only a few pigeons preserve the mark in the adult feathers; the greater number lose 

 it with the first molt and exhibit it as a mark in almost every stage of decline in 

 the juvenal feathers. I may add also that the pigeons share this character with a 

 very large number of birds, probably with the avian phylum as a whole. In other 

 birds the character is also sometimes permanent, but generally transitory, and fre- 

 quently reduced to vestiges that are scarcely more than just visible in the first 

 feathers. This interesting vanishing stage is seen, for example, in the young robin. 



These apical marks are very obscure or nearly wanting in the young of the more 

 highly bred fancy races, especially in those of a pale-gray color. They are best 

 defined in dark-chequered or blackish birds of the Columba affinis type. The mark 

 is crescentic in form, 0.5 mm. to 1.5 mm. wide at the shaft-line in the larger coverts. 

 As these crescents edge the tips of the feathers arranged in transverse rows across 

 the wing they produce a more or less obscure " lacing" effect, but as they stand alone 

 on each feather they do not suggest cross-bars. It is only very rarely that we find 

 in juvenal pigeons anything that would pass as cross-barring, 3 as this term is com- 

 monly understood, and it is only in such exceptional cases that the apical crescents 

 could be confounded with bars. 



I was for some time puzzled to find any meaning for this transient character. 

 I soon noticed, however, that it was invariably narrower in the two-barred than in 

 the chequered type of rock-pigeons, and further, that its vanishing-limit was reached 

 in races that had made the widest departures from the ancestral rock-pigeons. As 

 I carried the search for this character among the wild species of pigeons, I found 

 that the same rule generally held true, the mark tending to a vanishing-limit in the 

 higher species, although with some exceptions. 



As I had already discovered a tendency prevailing among both wild and tame 

 pigeons to lose spots and chequers, I began to realize that the apical crescent was a 

 thing of history and destiny; and that, in the case of the entire group of rock- 

 pigeons and their domestic descendants, it was a character slowly verging to com- 

 plete disappearance. 



But in a few of the "exceptions" just alluded to I found this character in what 

 seemed to be a primitive and typical condition. The oriental turtle-dove (Turtnr 

 orientalis) of Japan and China and its two nearer allies, T. ferrago of India and T. 

 turtur of Europe, are good examples. In these doves the apical mark appears at 

 full width in the juvenal plumage and persists practically unchanged in the adult 

 bird. Herein these turtle-doves betray a phyletic secret of no doubtful significance 

 when supplemented by another presently to be mentioned. 



1 Even such a transient color detail as the apical mark, which at first sight seems too trivial to deserve more than 

 mere mention, may have a wholly unsuspected significance when its genetic and phyletic relations are discovered. 

 If these marks have any significance beyond that of giving the plumage a "barred" appearance, we can only hope to 

 find it by tracing their history in the species, and by a comparative study of other species having a similar character. 

 The study of single characters from this point of view becomes a study in the evolution of species, and should give 

 us test-cases for theories far more conclusive than any furnished by sport-hunters and compilers of reported discon- 

 tinuities, and far beyond the depth of pretentious statistical curves, in which a linear series of multifariously collected 

 units is expected to reveal the secrets of an assumed multifarious variation. The artifice reveals the assumption 

 necessarily. No disrespect for the curve, for the curve tells the truth even when it gives us back the lie we put into it. 

 It is genetic history that gives us the sequence of variation— the true linear series. Such history can not be read in 

 the "large-number" curve. Comparative research and experimental culture are what the problem demands. 



• "Fundamental bars," as I have called them, are always present; but these are so obscure that they have been 

 universally overlooked, and hence may be left out of consideration here. These have been studied recently by Mr. 

 Riddle (Biological Bulletin, xn, 3, page 1C5, Feb. 1907). (These are figured and described in Chapter VIII.— Ed.) 

 13 



