PHrL0(5ENY OF THE PAL.llOGNATH^ AND NEOGNATH^. 249 



degree of divergence of type in the alimentary canal goes, the Ratites deserve their 

 accepted place at the bottom of the avian scale." 



I have been enabled to examine the intestines of all the Palmgnathce, save 

 Sfruthio. So far as my material goes it entirely confirms the observations of Gadow, 

 Mitcliell, and Beddard. 



Casuarius, Dromaus, and Apteryx all resemble one another rather closely. 



In Casuarius and Droimms the duodenal loop is wide, in Apteryx narrow. In 

 Casuarius, as Mr. Mitchell points out, the pancreatic and liepatic ducts open into a 

 diverticulum of the di.stal loop. In Dromaus and Aptenjx the duodenum is longer and 

 narrower, and tlie pancreatic and hepatic ducts ojien separately into the duodenum. 

 In other genera the small intestine is relatively sliort, and the convolutions therefore 

 are simple. The rectum is short and straight; in Casnarins and 7>romr*'?fs its inner 

 lining is thrown into a number of strongly marked irregular rugous folds. 



In Casuarius and Dromaus the ca^ca are relatively short, of a uniform thickness 

 throughout, and have a narrow lumen. In Apteryx the caeca are relatively much 

 longer, exceeding inches in length, whereas in the two former they do not exceed 

 5 inches. Furtliermore, in Apteryx the terminal half becomes suddenly and markedly 

 dilated ; each caecum thus presents a sausage-sliaped free end, borne on a narrow tube- 

 like stem. 



There are no valvular appendages in the caeca of tliese three forms. 



lihea is intermediate between Casuarius on the one hand and Sfruf/uo on the other. 

 The duodenal loop is narrow and closed. The convolutions of the small intestine 

 resemble Casuarius, the posterior rectal portion Strutliio. The pancreatic and hepatic 

 ducts open separately. The resemblance to Struthio consists in the relatively o-reat 

 length of the rectum ; but it differs therefrom in that it is only thrown into a single, 

 and not a series of complex loops. 



The cd'ca are relatively of enormous length, exceeding those of any other Pahvo- 

 (fnatJte. They a))pear, however, to vary in this respect very considerably. In the 

 series of measurements given by Remoucliamps [85] the variations in li. mavro- 

 rliyncha range between 1 foot 4.} inches and 2 feet 9 inches ; in R. americana between 

 2 feet 4 inches and 4 feet 8 inches. In a female of this latter species dissected by me 

 they attained a length of 5 feet o inches, accompanying the small intestine throughout 

 the greater pari of its length. The lumen was sacculated to within a short distance 

 of the extreme end of the caecum by means of a series of semilunar valves alternatino- 

 one with another. The position of these valves was indicated on tlie outside of the 

 caecum by means of a series of constrictions (fig. 7, p. 2-50). 



The intestinal coils of Struthio I have not had the good fortune to examine. I 

 therefore append Mr. Mitchell's description [60]. He writes: — "The first two parts 

 of the intestine and the three great veins were according to ty])e, the yolk-sac 

 diverticulum being conspicuous in the circular loop opposite the end of the median 



