246 DR. p. CHALMERS MITCHELL OX THE 



lengtliens, its closely bunched set of minor loops, developed 

 chiefly on the proximal limb of the pendant loop, must push 

 their way towards the middle line dorsal to the distal loop, 

 which in primitive mammals hangs nioi-e freely down in the 

 gut. 



Taking only the characters presented by the gut-pattei/ns as a 

 basis, it appears that the most primitive or generalized type had 

 a duodenum not well separated from Meckel's tract, Meckel's 

 tract, consisting of minor loops developed along the proximal 

 limb of the pendaiit loop, up to about the apex where the yolk- 

 sac was attached, a moderatel}^ straight recurrent limb baaiing 

 towards the middle of its length a functional ctecum (or more 

 probably a pair of functional cseea), a shai'ply bent colic flexure 

 close to the duodenum, where the pendant loop passed into the 

 rectal portion of the hind-gut ; that lectal portion considerably 

 longer than the length that it had to traverse, and thrown into 

 specially long minor loops at its proximal extremity. Apart 

 from rotation, the gut was suspended on a continuous primitive 

 mesentery, and the blood-vessels supplied the regions of the gut 

 to which they belonged. Changes from this primitive condition 

 occur in two directions : the pattern may be secondarily re- 

 duced and become even more simple, or it may become more 

 elaborate. 



When cha,racters are used for the purpose of classification, it 

 may be convenient, in the absence of other information, to place 

 creatures in the same group because they have retained ancestral 

 conditions, but if the classification is intended to state the degree 

 of aftinity, then it must be remembered that thei'e is no a j^riori 

 reason to suppose that amongst the descendants of a common 

 ancestor, the groups that have retained an ancestral character 

 are more closely related than the groups that have lost it. On 

 the other hand, the common possession of a well-marked elabora- 

 tion of the primitive type appears to present some ground for 

 implying aftinit}^ 



As in my summary of 1905, I associate the Marsupialia, 

 Xenarthi'a,, and Tubulidentata as displaying, on the whole, the 

 most ancestral type of gut-pattern, with the proviso that this 

 association does not imply close affinity but mei'ely the retention 

 of a common inheritance. I think it safer to' exclude the Galeo- 

 pithecidie from this association, as my information with regard 

 to that Order is second-hand. I note with regard to the Mar- 

 supials, however, that they contain tAvo dejiartures from the 

 ancestral type. In some of the small Polypi'otodonts the gvit- 

 pattern is extremely reduced, with complete loss of cpeca and 

 obliteration of clear distinction between the different regions. In 

 other Marsupials, such as the Phascolarctidje, the hind-gat has 

 attained an elaboration recalling that of higher types. In the gut, 

 as in many other parts of their structure, the Marsupials appear 

 to forecast, on a lower level, and in a, more fluctuating condition, 

 clahorations that liocojiic doHuitc and '• lixed " in higher types. 



