Vol. Xixn Snodgrass, Food and Size of Bill in Geospiza. 379 



1902. J 



mens taken in April on James island had fed as follows ; in one 

 stomach were seeds Nos. 11, 26 and 28 ; in two others seeds Nos. 

 18 and 26 ; in another seeds Nos. 57 and 59 ; in the fifth seeds 

 Nos. 14, 15, 18 and 59. Of two taken in June on Bmdloe one 

 had in its stomach only seed No. 46, the other only seed No. 44- 

 These very scant data would seem to indicate that Geospiza strenua 

 uses but little selection in the choice of its food. Altogether it has 

 been found to eat seeds Nos. 11, 14, i5^ 18, 22, 26, 28, 44, 46, 57 

 and 59 The James specimens alone had eaten seeds Nos. n, 

 14, 15, 18, 26, 28, 57 and 59- This list is somewhat different 

 from the diet of six specimens of G. scandens fatigata taken at the 

 same time on James. The stomachs of these birds gave the fol- 

 lowing: seed No. 2 in one bird, No. 14 in two birds, No. 15 in 

 three birds, No. 18 in one bird, No. 21 in one bird, No. 26 in one 

 bird No. 41 in four birds, No. 59 i^ one bird. The species of 

 seeds forming the list in the two cases are almost the same, the 

 main difference is in the proportions of the seeds present. It is 

 a question whether the evidence in this case should not be given 

 to proposition III. It, however, shows the weakness of the proof 

 on which proposition V could be stated as a general fact. 



A -ood example of the proposition under consideration may be 

 derived from a comparison of the food of Geospiza Jortis (PI. XI, 

 Fig 3) at Tagus Cove, Albemarle, with that of G.fuhgmosa par- 

 vula at the same locality. As has already been shown, the food 

 of nineteen Tagus Cove specimens of the latter species was as 

 follows: seed No. i had been eaten by two birds, seed No. 2 by 

 one bird, seed No. 4 by two birds, seed No. 6 by two birds, seed 

 No 8 by nine birds, seed No. 15 by one bird, seed No. 22 by five 

 birds seed No. 24 by three birds, seed No. 41 by two birds, seed 

 No 48 by one bird. The table shows the following composition 

 of the food of thirteen Tagus Cove specimens of Geospiza fortts 

 taken also during January. Seed No. 2 had been eaten by one 

 bird seed No. 3 by four birds, seed No. 4 by one bird, seed No. 

 8 by one bird, seed No. 9 by one bird, seed No. 22 by five birds, 

 seed No. 28 by two birds, seed No. zi by one bird, seed No. 35 

 bv three birds. The only important difference in these two cases 

 is' the predominance of seed No. 8 in the food of G.fuUginosa and 

 Its scarcit)- in that of G. fortis. These two species have somewhat 



