1901.] EVOLUTION OF PATTEKN EH FEATHEBS. 319 



The distinctive patterns of birds caused by different patterns on 

 a number of feathers taken together owe their origin to other 

 causes, and must not be taken into account in this paper except in 

 so far as they have been directly modified from these primitive 

 patterns. This class of markings, although the most important 

 for the bird's welfare, is probably the most unstable, being con- 

 tinually altered to keep the race distinct when it meets with various 

 other nearly allied members of its genus, either in extending its 

 range, or from some other cause. It is worth mentioning on this 

 subject that when a race finds it essential to differentiate itself 

 from a closely allied form, it has to cover up its distinctive mark as 

 well as to evolve a new one ; and consequently we may suppose 

 that some of the markings on birds are the degenerate race-marks 

 of former generations. The genus Fringilla offers a good example 

 of this. 



There are eight species or forms known, namely : — 



F. teyclea inhabiting Teneriffe. 



F. Calebs „ Europe generally. 



F. maderensis „ Madeira. 



F. maderensis moreleti . . „ Azores. 



F. maderensis canariensis „ Teneriffe. 



F. maderensis palma .... „ Las Palmas, Canaries. 



F. spodAogenys „ Algeria. 



F. montifringilla „ Europe generally. 



The whole of these species (with one exception, F. teydea) may 

 be recognized by two distinct cross-bars on the wing, especially 

 noticeable when the bird is flying ; these cross-bars being formed 

 by a more or less broad white tip to certain of the wing-coverts. 



In only two places is more than one species of this genus found, 

 namely, in Europe, where we find F. coelebs and F. montifringilla, 

 the latter having, however, a different breeding range ; and at 

 Teneriffe, where there is a very restricted forest-form, F. teydea, and 

 a species of the more normal type, F. canariensis. 



It is instructive therefore to notice that in these two cases only, 

 where the risk of intermingling between the species might occur, 

 do we find any attempt on the part of one or other of them 

 to alter or modify the characteristic cross-bars of the wing. 

 F. teydea accomplishes this more completely and successfully than 

 F. montifringilla ; but, considering its restricted range and the com- 

 paratively greater amount of in-breeding which must go on, such a 

 result is only natural. F. teydea obliterates the cross-bars altogether, 

 and does not assume any other distinctive mark, the eontined 

 space of its range tending to render such a mark unnecessary. 



With /•'. montifringilla the case is different, ami while the 

 cross-bare are being obliterated, it has assumed a further distinctive 

 marking in having a white rump. We are, however, I think, 

 quite justified in Coming to the conclusion that this species is still 



in a state of transition, since it is seldom (hat the rump lb purely 

 white, being frequently interspersed with darker feathers, while 



