48 PROCEEDINGS OF THE \ ITIONAL MUSEl I/. vol. 



. si 



to attend to them. The next day I found that, owing to their having 

 been placed near some steam pipes, the geese were too far gone for 

 preservation, and I was reluctantly obliged to throw them away. I 

 took color notes and measurements of each specimen, however, which 

 may prove of interest. 



No. 1. Dimensions: Wing 14.90 inches; tail 5.75 inches; rectrices 

 16. White cheek patches confluent on throat, with a few scattered 

 dark feathers in the median line; white collar about lower neck j 

 continuous all around, but only about half an inch wide; white of 

 anal region sharply marked oil' from brown of breast and abdomen. 



No. '_'. Wing 15.75 inches; tail 5.62 inches; rectrices 16; white 

 cheek patches divided for their anterior half by a black peninsula 

 three-quarters of an inch wide at- the base, narrowing to five-eighths 

 of an inch at its end, and a row of scattered black specks from the 

 end of the peninsula to the posterior border of the cheek patches in 

 the median line of the throat; collar on lower neck 1 inch wide, 

 but narrowing toward dorsal surface, where it is crossed by a bar of 

 black, one-quarter of an inch in width, in the median line; under- 

 pays as in No. 1. 



No. 3. Wing 16.12 inches; tail 5 inches; rectrices 14; cheek 

 patches divided for their anterior half as in No. 2; white collar on 

 lower neck three-quarters of an inch wide in front, decreasing in 

 widtli toward dorsal surface, where it is crossed by a bar of black; 

 underparts as in Nos. 1 and 2. 



No. 4. Wing L5.25 inches: tail 5.50 inches; rectrices 16; cheek 

 patches small, and entirely separated by an isthmus three-quarters of 

 an inch wide at the anterior and posterior ends, narrowing to one-half 

 an inch in the middle; white collar on lower neck one-quarter of an 

 inch wide in front, narrowing behind, where it is much broken up 

 and divided by a black bar in the median line; upper surface some- 

 what darker than the lower, the other birds having been uniformly 

 colored; t his specimen was also slightly darker throughout , and was 

 apparently brooding, a Large patch on the breast being bare of feathers. 



In a patch of long grass near the top of a bluff just behind the 

 beach I found a nesl containing three eggs belonging to this species. 



At Attn I found tin- goose not uncommon, but the} do not breed 

 there. Several were seen on a small islet to the right of the harbor 

 entrance 1 , and others Hying overhead at various places in the low- 

 land-. The natives here told me of their breeding in great abundance 

 on Agattu, and also in Lesser numbers at Semichi, and they also said 

 that the ravens on those islands steal their eggs and bury them for 

 winter consumption. The absence of foxes from Agattu and Semi- 

 chi undoubtedly accounts for the occurrence of this species on these 

 islands in such abundance. 



