INTESTINAL TRACT OF MAMMALS. 207 



fig. 2). In these drawings the position of the cBecum on the 

 i-ight side is well shown. Dr. Beddard's own generalised dia- 

 grams of the mammalian gut (Beddard, 1908, text-figs. 122 & 

 123) show the same point. Now, if Dr. Beddard's own diagram 

 of the alimentary tract of H. capensis (Beddard, 1908, text- 

 fig. 115) be examined, it will be seen that he represents (and all 

 my observations confirm him on this point) the paired ceeca in 

 the undisturbed condition as lying on the right side, in the true 

 position of the normal mammalian caecum, with which he does 

 not homologise them, and the unpaired caecum as attached to the 

 gut nearly in the middle line, much to the left of the paired 

 cseca, and therefore in a position in which the normal mammalian 

 csecum never lies. As a matter of fact, the accessory caecum 

 of the Hyracoidea, both in the young and the adult, lies in 

 a region of the body-cavity always occupied in mammals by the 

 coils of the small intestine. 



Although Dr. Beddard (1908, p. 595) makes the general state- 

 ment that the series of facts (rotation of the gut in the body- 

 cavity, mesenterial attachments, formation of " fixed " loops) 

 cannot yield any accurate classificatory results, he appears to 

 rely on precisely such facts in his endeavour to show that the 

 unjjaired cfecum of Hyrax is homologous with the normal caecuin 

 of mammals, and that the intestinal tract of the Hyracoidea is 

 to be associated with that of the Perissodactyle Ungulates. As 

 he himself has shown conclusively, rotation of the gut occurs in 

 almost every group of mammals, and therefore its presence, or 

 even the stage to which it has reached, does not assist us in the 

 attempt to detect relationships. I have already (suiwa^ p. 184) 

 shown that it is necessary to distinguish carefully (a point that 

 Dr. Beddard has overlooked) between the secondary connections 

 and the primitive mesentery, as the former are almost certainly 

 convergent adaptations. Even assuming, however, that the 

 ligaments might yield evidence of affinity, those that are present 

 in the Hyracoidea do not support Dr. Beddard's argument. A 

 strong wide ligament attaches the unpaired caecum to the portion 

 of the gut which leaves the caecum. This is more extensive in 

 H. capensis (Beddard, 1908, text-fig. 115, 1) than in D. dorsalis, 

 in which it extends no further than the additional small csecal 

 pouch (text-fig. 12, C. 3) present in that species. Dr. Beddard, 

 in directing attention to this, points out that the single caecum 

 of mammals, however small, is usually, possibly invariably, 

 attached to the adjacent wall of the gut by such a ligament. It 

 happens, however, that the mesentery of the true caecum in other 

 mammals passes between the true caecum and the ileum, that is 

 to say, the portion of the gut entering, not leaving the caecum. 

 I do not know of any exception to this relationship, which is in 

 correspondence with the appearance that the caecum pi-esents of 

 ■ being an anteriorly directed outgrowth of the hind-gut, running 

 forwards roughly parallel with the ileum. This normal mesentery, 

 stretching between the caecu.m and the ileum, is absent in the 



