1908.] AJSTTEOHINOMYS AND OTHER MAKSUPIALS. 595 



But other genera, e. g. Mas and Arvicanthis, with one ansa, 

 jjaracsecalis, and Sciums with two ansae, dextra and sinisti'a, and 

 Aulacodus with only the ansa coli dextra, render it impossible to 

 make hard and fast lines of division. 



It will be noticed from the above account of the several stages 

 seen in the Mammalian gut, that the Lemurs are the only group 

 in which every stage but one is to be seen in a well-marked fashion. 

 This fact of itself is enough to negative any accurate classificatory 

 results to be deduced from the series of facts brought together in 

 the present communication, though I offer later (p. 596) some 

 observations upon the affinities of different groups as judged by 

 the varied modifications of the intestinal canal. 



Furthermore, it will be gathered fi'om what has been brought 

 forward in the present communication that the Marsupials, 

 although their intestinal tract shows in a well-developed fashion 

 only two stages, show indications as it were of the third stage. 

 Occasionally an ansa coli exists, while secondary connections 

 between the duodenum and commencement of the colon are found 

 in Trichosurus. Arising, as is now believed, from some early 

 Eutherian tj^De, the Marsupials seem to have retained the poten- 

 tialities of intestinal development exhibited in the later Eutheria. 



It is important moreover to note that the attachment of the 

 omentum to the colon in Trichositriis is to the extreme right of 

 the latter ; for it is on this side that the attachment commences 

 in Man (Johannes Miiller quoted by Klaatsch). 



Simplification of structure does not always imply an ar-chaic posi- 

 tion with reference to allied forms which show a less simple anatomy. 

 It has been again and again pointed out that size is an element 

 which is not to be left out of consideration in weighing such 

 apparently archaic structural conditions. Foi- example, the smooth 

 brain of many small mammals is not to be interpreted as evidence 

 of the lowly position in the series of such smooth-brained types. 

 On the other hand, the simple organisation of a ISTaiid Oligochaete 

 as compared with that of a large earthworm, may be at least 

 partly interpreted as simplicity not altogether due to reduced 

 size. It is important therefore to note that the simple intestine 

 of Antechinomys ?i\wa.g upon a single continuous mesentery is not 

 a feature confined only to such small mammals. In comparing 

 this form with Arvicanthis, which is even smaller, we find in the 

 latter the usual rotated intestine with even a fixed colic loop. 

 And other examples will be apparent from the foregoing pages as 

 well as from previous writings on the subject. At the same time it 

 is not to be forgotten that other instances may be due to simpli- 

 fication, and to belong therefore to a different category though 

 apparently quite similai*. 



Although it is true that the specialised loops of the colon are 

 often associated with a colon of great length, there is as it would 

 appear no necessary connection between length and complexity, 

 or even occurrence, of these special loops. The existence of such 

 coils is in fact a character of given orders of mammals. Among 



