ORNITHOLOGICAL CLUB. 



61 



diflferent from that heretofore considered to embrace the most im- 

 portant characters. In the case of Carpodacus, there has been 

 reached in the Guadalupe race almost, if not quite, that degree of 

 diilerentiation which distinguishes the Galapagoan genus Geosjyiza 

 from continental Guiraca, the modifications being moreover of pre- 

 cisely the same character. Junco insularis may likewise be com- 

 pared with the species of Cactornis, in which the bill has be- 

 come so extremely produced as to have almost lost its fringilline 

 character. 



The three species above-mentioned exhibit the local modifications 

 to the greatest extent, but the rule may be traced through the 

 whole series ; and to show the exact extent of these modifications 

 of form we present the following table of measurements of each of 

 the Guadalupe species compared with those of the mainland repre- 

 sentative form, the measurements representing the maxinmm and 

 minimum of a large series of each : — 



* The Guadalupe species are in Italics. 



A close perusal of the above figures leads to the discovery of 

 some exceptions to the rule of variation in proportions. Thus, 

 Regxdus obscurus, while conforming in other respects, does not have 

 the tail constantly shorter than R. calendula, although it averages 

 shorter. There is also no appreciable difference in the absolute 

 length of the tarsus in the two forms of Thryomanes and Salpinctes, 

 though the compar^ative difference in favor of T. brevicauda and *S'. 

 obsoletus, when contrasted with the length of the wings and tail, is 

 very marked. 



