1908.] ANTECHINOMYS AND OTHER MARSUPIALS. 583 



immediately aftei- it has left the stomach to the colon opposite 

 to the attachment of the first part of the omentum. It may be 

 that the attachments respectively to the colon immediately after 

 it has left the paired cteca, is an argument in favour of regarding 

 those cajca as the equivalents of the unpaired cfeca in other forms. 

 For in Rodents the attachments of the membianes in question 

 are sometimes to the colon immediately after it has emerged 

 from the caecum in those animals, and the same relations are to be 

 found in Lemurs. But against this i-esemblance may be placed 

 the facts of the attachment of the membranes in question in 

 Dasyprocta. In that Rodent, as I point out, the omentum and 

 the duodenum are inserted upon the ansa coli, which lies at 

 a considerable distance from the point of emergence of the colon 

 from the caecum. The evidence therefore cannot be regarded 

 as very strong. Whereas the evidence already dealt with against 

 the identification is very strong. 



It will be clear from the foregoing that Dr. Mitchell's statement 

 that " the hind gut divided into a simple colon and rectum merely 

 conforms to the fundamental mammalian plan " is not coi'rect. 

 Nor can I agree Avith him in the further observation that " the 

 general pattern of the intestinal tract in Hyrax, however, sug- 

 gests no affinity with the patterns exhibited by Rodents and 

 Ungulates." It appears to me to resemble both. But this is 

 of course a matter of opinion. It is not without importance to 

 observe that Hyrax shows some variation in certain of the 

 features described above. Since writing the account which I 

 have just given of this Ungulate I have dissected three other 

 specimens, all of small size like that from which the above 

 account has been practically entirely di-awn. In one of them, 

 the colon at the end of the transverse section at the further 

 attachment of the omentum is fixed into a short ansa coli which 

 is not very narrow, i. e., the two limbs are not closely approxi- 

 mated. This corresponds in position to the splenic flexure of 

 human anatomy, and is very like the ansa coli sinistra of Tamias 

 st7'iata described under my account of the anste coli of Rodents. 

 The two other specimens did not show this specialised loop. 

 Furthermore, the mesocolon undergoes some vai-iations in its 

 i-egion of attachment to the doi'sal parietes. I did not observe 

 the exact arrangement in the first example dissected. But in 

 the three now under consideration there were three difterent 

 modes of attachment, thus cleai-ly showing a great variation. 

 In the individual just i-ef erred to, this mesentery was attached 

 altogether oiitside of the left kidney. In a second individual, the 

 left kidney lay for the gi^eater part to the left of, i. e. outside of, the 

 mesocolon, but the attachment of the latter was in part to the 

 kidney and cut off an angle of that viscus anteriorly and to the 

 right. In the thii-d example, the line of attachment of the meso- 

 colon divided the kidney into two longitudinal areas, of which 

 the inner lay within the mesocolon area and the outer lay outside 

 of the mesocolon. The pocket of peritoneum referred to was 



