2 20 Lucas, Bird Skeletons frotn Guadalufe Island. [April 



while chanjj^e from any inherent tendency of a species to vary 

 seems to come about very slowly and require a vast stretch of 

 time for its accomplishment. 



Comparison of Plpilo consobrinus with Plpllo mactilatus 

 mecralonyx and P. erythrophthahiuis shows a considerable fall- 

 ing ofl'of the island bird in the length of the sternum, for while 

 the legs and wings of all three species are practically alike the 

 sternum of consobrinus is but little more than two thirds as long 

 as that of erythrophthahnus. Pipilo erythrophthalmns is 

 much the strongest of the three species in its wing, for although 

 the wing itself is but a trifle longer than in the others the sternum 

 is not only longer, but deeper than in either consobrinus or mcg- 

 aloiiyx, indicating well developed pectoral muscles. As all the 

 Chewinks spend much of their time upon the ground, similarity 

 of habit in this respect would naturally account for similarity in 

 the size of the leg. The migrations of Pipilo tnaculatns inega- 

 lonyx^ are short, this southwestern species inhabiting the moun- 

 tains in summer and descending to the valleys for the winter. The 

 migrations of Pipilo erythrophthalmns on the other hand are 

 extensive, and its greater sternal development is simply a residt 

 of the greater length of its travels, while the restriction of Pipilo 

 consobrinus to one locality, coupled with its ground-loving 

 habits, has brought about -the diminution of its flying apparatus. 



Jiinco i??sularis^ when compared with y. hyemalis^ not only 

 shows great sternal reduction, but reduction in the length of the 

 wing, although the humeri of the two species are much alike. 



The case of these birds parallels that of the Pipilos, Junco 

 hye?}ialis being a bird of extensive range and consequently good 

 powers of flight, while insular is is of restricted range and equally 

 restricted flight. 



Carpodacus aniphis is well named, for it is a stout, well- 

 rounded bird, slightly larger than Carpodacus cassini^ and al- 

 most twice the bulk of its nearer relative Carpodacus frofdalis. 



Comparison shows that C. amplus is ahead of both these in 

 length of leg, and that it leads C. frontalis in length of wing, al- 

 though showing some falling oft" in the length of the sternum. 



All in all the island bird seems to have undergone but little 

 change from its restricted habitat, and if it has lost in wing power, 

 this has been compensated for by increase in the length of leg 

 and size of skull, this exceeding that of C. cassini. 



