1908.] ANTECHINOMYS AND OTHER MARSUPIALS. 589 



condition being therefore much as in the half -grown Hydrochoerus 

 just referred to. 



Tullberg's figure of the Agouti is not quite so satisfactory as 

 are the majority of his figures. It is so small that various details 

 are left out. I therefore venture to supplement him by another 

 figure of an allied species Dasyprocta punctata (text-fig. 11 9 j. In 

 this specimen the colic loop is, as shown, rather larger than in 

 D. aguti. It is precisely as in the Lemurs Galago and jVycticebios 

 (see p. 578). The first part of the colon, as correctly shown by 

 Tullberg, runs parallel with and very close to the caecum, to which 

 it is attached by a mesentery. There is a kind of attempt — so to 

 speak — at its origin of an ansa paracsgcalis like that of the Murines. 

 The duodenum is attached both to the colic spiral and to the 

 commencement of the caecum, which bends back upon itself at 

 its free end as shown in the figure. The great omentum is 

 attached to the colon where it emerges from the spiral and also 

 to the contiguous part of the spiral itself. It is not, however, 

 attached to the whole left border of the spiral as in the Lemurs 

 mentioned. I should mention that the cseco-colic ligament is 

 iittached along one of the two muscular bands upon the cfecum, 

 the other being on the opposite side. 



The small Barbary Mouse, Arvicanthis picmilio, has the 

 simplest colon of any Rodent which I have had the opportunity 

 of examining *, and the conditions characterising this genus have 

 not been dealt with by Tullberg. The colon itself is relatively 

 short and thrown into no temporary folds. The caecum lies on 

 the right side of the body rather low down, and the colon ascends, 

 shows a transverse region, and then forms the descending colon. 

 There is only one ansa coli present, and that is just where the 

 colon emerges from the caecum. It is there twisted into a short 

 spiral. This ansa parac»calis is in principle like that of other 

 Rodents such as Cricetus. But it is the only loop present in Ariri- 

 canthis. It is noticeable that it has the characteristically Murine 

 form. I could find no trace, at the angles formed by the bending 

 of the colon, between the transverse and descending regions, 

 of even so rudimentary a persistent loop as there is in Cricetus. 



Tullberg has described various points in the anatomy of the two 

 species of Otomys, viz. 0. icnisidcatus and 0. bisulcatus, but has 

 not dealt with the gut except to give the proportions of the 

 several regions in the latter of those two species. Having had the 

 opportunity of dissecting 0. iri-oratus, I am able to fill in that 

 lacuna in our knowledge of the Rodentia. In this Rodent we 

 find almost exactly the same characters as in Mu,s. That is, there 

 is only one colic loop and that is situated just at the commence- 

 ment of the colon where it emerges from the caecum. This ansa 

 coli is doubled upon itself once, and this forms an " N " which is 

 bound down to the caecum. There are no other ansae along the 

 course of the colon. The great omentum seems to be not present 



* Mus rattus is equally simple, and like M. decumanus figured hy Mitchell — with 

 perhaps even a less marked ansa coli. 



