AND MYOLOGY OF THE MAKSUPIALIA. 863 



In Petaurus sciureus (text-fig. 75 E) the globular fundus is 

 directed forwards and to the left. The duodenum, whose com- 

 mencement is notched, leaves its dorsal aspect. In P. breviceps 

 (text-fig. 75 D) the duodenum has no notches, but a groove 

 marks the pyloro-duodenal junction. The greater curvature is 

 sacculated. , 



In Dendrolagus ursinus (text-fig. 75 A)* the greater part is 

 concealed by the liver, the only parts visible being the cardiac 

 cul-de-sac and greater curvature. The foi-m and general cha- 

 racters resemble those of other species already described by 

 Beddard (1), Owen (12), and Oarlsson (2). These authors com- 

 pared the stomachs with those of other Macropodidae, and 

 described the extent of the glandular and non-glandular parts of 

 the mucosa.. 



Owen and Beddard described groups of follicular glands lying 

 along the sides of the central tract (a) in D. mustios and 

 Z>. hennetti. In D. ursinihs they are replaced by long glandular 

 strips (c), into which branches of the abdominal parts of the 

 vagus nerves can be traced, and the vagus branches probably 

 contain sympathetic fibres. From the examination of the 

 stomachs of animals belonging to iaiost of the mammalian orders, 

 I have come to the conclusion that the branches of the vagi are 

 greatly increased in numbers when there is a special gastric 

 glandular apparatus. These branches pass directly to the glands 

 as in D. ttrsiiius, or run through a plexus with ganglia as in 

 Phascola7-ctos cinereus (15). The pylorus does not project into 

 the duodenum in Dendrolagus. 



Osgood described a special gastric gland in Ccenolestes obsciorus, 

 but did not mention its nerve supply. 



The histology of the stomachs of the Marsupialia has already 

 received considerable attention, and Oppel (8) has collected and 

 analysed the various papers. I have nothing new to add to 

 these accounts. Nematode worms are present in many stomachs. 



The Intestinal Tract : — The observations recorded here are 

 supplementary to those of Chalmers Mitchell (6, 7), Cunningham 

 (3), Owen (10), and Flower (20). 



In MetacMrus ojoossimn the well-marked duodenal Ibop is wide, 

 thin-walled, and extends posteriorly for a considerable distance. 

 The flexure between its second and third parts is connected 

 by the usual peritoneal ligament to the abdominal parietes and 

 meso-colon, but is not continuous with the splenic ligament. 

 The caecum is two and a half inches long. It is capacious, and 

 its interior is devoid of folds and septa. Meckel's Tract and the 

 large intestine differ in no respect from that of Didelphys 

 virginiana (6). The caecal mesenteiy has a concave free edge 

 and the angle between the caecum and Meckel's Tract is bisected 

 by an artery (text-fig. 76 A). 



The caecum of Perameles ohesula (text-fig. 76 B) is one and 



* The specimen was specially hardened by injections of forniiiliu to show the 

 topography of the stomach. 



