INTESTINAL TRACT OF MAMMALS. 229 



to Marsupials is not so close as to Monotremes. This is a point 

 to which I had paid no special attention in 1905. Dr. Broom's 

 conclusions are derived from inv^estigation of the organ of 

 Jacobson, and certainly the conditions that he has found appear 

 to form a broader basis for S3^stematic conclusions. I a,m not 

 quite certain, however, as to whether or no he means to imply 

 that the " Ccenrhinate " type of organ, which he finds to occur 

 in Tal])a, fSorex, Erinaceus, Gymnura, Centetes, and other 

 normal Insectivora, as in Carnivora and Ungulata and most 

 higlier mammals, is a derivative of the more ancestral " Archfeo- 

 rhinate " type which he has found in l\ipaia, 3Iacroscelnles, 

 Diprotodonts, etc. If he means that the Archa?orhinate type is 

 the primitive type, and therefore to have been present in the 

 common stock, the fact that it has been retained by certain 

 forms is no conclusive argument for placing these forms together. 

 As he finds it to occur in Monotremes, on the one hand, and 

 in Dasypus, Orycterojius, and Rodents on the other, I suspect 

 that it is, like the presence of a, primitive gut-pattern, a character 

 tliat must be used with caution in classification. 



Without carrjdng further this question of breaking up the 

 Insectivora, I may sum up by saying that the gut-patterns of 

 the group start from an extremely simple type, and show 

 successive stages of secondary simplification. 



Order Ciiikoptera. Rhhiopoma microphyllum (text-fig. 24). 

 Artiheus jdanirostris (text-fig. 25). 



Since I wrote in 1905 I have had tlie opportunity of examining 

 the intestinal tract in some other Bats, of which the most 

 interesting was an example of Bhinopoma microphyllum ( = 

 R. hardwickii). The latter and Megaderma spasma were the 

 two Bats in which Owen found a ctecum present (Owen, 1868, 

 p. 429). In Rh'mopoma the duodenal region is well separated 

 from Meckel's tract. Meckel's trnct makes up the greater 

 portion of the gut, and consists of a number of veiy^ ii'regidar 

 minor loops, arranged so that they nearly complete the circum- 

 ference of a circular exjianse of mesentery, suspended hy a nari'ow 

 stalk to the' mesentery of the duodenum in front and to that of 

 the hind-gut posteriorly. In other words, the whole of the 

 recurrent limb of the pendant loop is occupied by Meckel's ti'act, 

 and it is only where it bends backwards to form the short and 

 nearly straight rectal portion that the attachment of the caecum 

 marks the transition from fore-gut to hind-gut. The caecum is 

 short and conical. The hind-gut may be regarded as without 

 a colon, but consisting mei'ely of a i-ectal portion. 



In the unexpanded condition, the subsidiaiy coils of Meckel's 

 tract are iri'egularly folded over the mesentery so that they 

 make up a large irregular mass visible as soon as the abdominal 

 cavity is opened. The duodeniun is also folded backwards, and 

 cannot be seen until the mass of the fore-gut has been pushed 



