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 arc 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 I he coverts \vcre t hen 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 lire- 

 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 r, ;XT .,,,,-, , fK 



many cases, no connecting phases are Pin-featben marking the region of tb oblkpie 

 normally presented. The appearances, <>inmon pip-on of 17 days. Tin- n^ion of the pins 



., . ., i/M ., is the region in which tin- feathen most nearly up- 



thercfore, seem to exactly fulfil the re- pro . M .\ t ,i, ( , j.,1,,1, ( . |,, r . 

 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 perfect ly 

 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. 



e rtna 



