37^ 



Correspondence. " t 



L Oct. 



where so many generalizations are made he could have been more judicious 

 in selecting an example to fit the term. In speaking of color determining 

 habit I referred only to the general shade of color — brown genera would 

 be forced by natural selection to the ground and olive-green birds to the 

 trees; but in speaking of the Brown Creeper I refer to its detailed 

 markings and streaks, in which we have not a perfect illustration, but 

 "the nearest approach" to an instance of special protective resemblance. 



In speaking of the Passenger Pigeon Mr. Allen takes the trouble to 

 italicize the assertion that the tail markings to which I allude as recog- 

 nition marks "are found only at the extreme base of the tail, within the 

 area normally concealed by the coverts, and are therefore not visible 

 under any ordinary conditions." If he will take the trouble, instead of 

 merely looking at a skin with closed tail, to spread t he tail feathers, as a 

 bird does at every turn in its evolutions, he will find a conspicuous broad 

 band of dark brown strongly relieved against the white of the under tail- 

 coverts and contrasted also with the conspicuous white outer tail-feathers. 

 Although I have seen the Passenger Pigeon alive I do not now remember 

 how distinctly the tail markings showed, but I have in no instance in the 

 text of my paper implied that the conclusions were based upon the study 

 of live birds, except in certain instances where this was stated. [6] Slips 

 in nomenclature are never pardonable in a work of this sort, but by way 

 of explanation I may state that I was suffering from an attack of nervous 

 prostration during the publication of the latter part of the work, and was 

 physically unable to give it the care which it demanded. 



Mr. Allen fails to see any use in the plate showing the evolution of the 

 pattern of head markings, since, as I have said in the text, the relationships 

 "are not supposed to be genetic." The plate is intended to show that 

 among living North American birds various types of head markings exist 

 which are related more or less nearly to one or all of the five types with 

 simple longitudinal streaks. There is no way in which we can now learn 

 the colors of extinct birds, and it is consequently entirely out of the ques- 

 tion to think of presenting a genetic series of head markings to show their 

 evolutionary sequence. The most that can be done is to show that living 

 birds happen to represent different stages in an ideal sequence from a 

 streaked plumage, and this taken in connection with the fact that the 

 streaked feather is the elementary type of feather marking, [7] and with 

 the a priori considerations as to why it should be so, serves to confirm, 

 without necessarily proving, the supposition that the head markings have 

 all been evolved from longitudinal streaks. 



The fact that in comparing low groups of birds like the Pigeons and 

 Tinamous with such high groups as the Thrushes and Sparrows the latter 

 are found to have a streaked plumage where the former have not, is in no 

 wise contradictory to the assertion that the streaked plumage is the 

 primitive type. Surely no one ever made the absurd assertion that color 

 development advanced from the lowest to the highest groups of birds pari 

 passu with structural development. How then could we explain the high 

 development of markings in Auklets and Ducks, and the brilliant plumage 



