258 J. L. Bonhote: 



" reversion " are merely variations, which happen to resemble 

 other species and cannot be considered as evidence that the 

 species so resembled are archaic forms, or that they are less 

 remote from the common ancestor from which all the Anatida 

 may be supposed to have originated. 



Some time ago in a paper to the Liiiiigean Society l I 

 attempted to show how the presence or absence of colour 

 tended almost invariably to make its appearance first of all 

 on certain definite tracts common to mammals and birds 

 alike, and for which I proposed the name " poecilomeres " 

 (ttoikolXo? /xepos=spotted part.) 



It may perhaps be worth while to briefly recapitulate a 

 few of the main characters of these poecilomeres so far as 

 they concern the class Aves. 



Poecilomeres are situated on the following parts — viz., 

 chilly, malar stripe, maxillary stripe, a spot above and slightly 

 in front of the e}*e, a spot below and slightly behind the eye, 

 the ear, crown of the head, occiput, fore-end of sternum, 

 vent, rump, thighs, wrist, and shotdders (above and below) . 



Now there is hardly any species of bird on which one or 

 more of these poecilomeres is not " picked out " (to use a 

 painter's expression) in some colour different from that of the 

 surrounding parts, and in fact most of the so-called recogni- 

 tion or protective markings will be found on these patches. 



On the other hand, among' many species the differentiation 

 of colour on the poecilomeres is not so conspicuous as to 

 attract the eye or to serve in any way for protection or 

 mimicry, yet we still find them marked by differences of 

 colour so slight that unless specially looked for they would 

 never be noticed. 



Or, again, some species occasionally, but not invariably, 

 show a few white feathers on certain parts of their body, and 

 when such is the case it will be found that these white 

 feathers appear on the poecilomeres. 



In a paper on the Hybridisation of Ducks it would be out 

 of place to dilate on poecilomeres at too great a length, 

 and since there is hardly a species in which examples of 

 these poecilomeres may not be found, I will content 



'Summary published Proc. Linn. Soc, Vol. XXIX., p. 185 (1904); more 

 detailed account in " Knowledge," December, 1905, January to April, 1906. 



