PROCEEDINGS OF THE NATIONAL MUSEUM. 



307 



between Certhiola and the Drepanididte, and the convohitions of 

 Hemignathus olivaceus very nearly coincide with those of C. caboti.* 



But in both groups there is varying complexity of convolution among 

 the dift'erent species, and in neither is there any adherence to a given 

 pattern. Among the Sandwich Islands birds there is, in the majority 



Fig. 8.— Intestinal convolutions of (1) Glossoptila campestris: (2) Coereba cyanea: (3) Dendroica 

 coionata; (4) Certhxola caboti; (4a) Certhiola cai>o(i, central portion; (5) Ccereba coervla; (6) Hemigna- 

 thus olivacei'S; (6a) Hemignathus olivaceiis, with coil opened out to show convolutions; (7) Acanthor- 

 hynchus tenuirostris. 



of specimens tigured, a slight peculiarity in the manner in which the 

 intestiue begins to uncoil from the center. When looking from below 

 at the right side of the viscera, the intestine is seen, roughly speaking, 

 to start from the stomach and in a decreasing spiral or series of loops 



* There is at first sight an apparent discrepancy between Dr. Gadow's figure and 

 that shown in fig. 8 (6), of this paper, but this is due to the fact that Dr. Gadow's 

 specimen has a longer and more closely twisted intestine, so that the point of rever- 

 sion is ditt'erent in the two. 



