192 DONALD WALTON DAVIS 



It will be seen that the full number of mesenteries (4) in the 

 half of a regenerating zone adjacent to a complete non-directive 

 bounding mesentery is found in 93 cases, the reduced number 

 (2) in 3 cases. On the side of old incomplete bounding mesen- 

 teries of the first grade the number of instances of the reduced 

 number (3 mesenteries) is greater (22 to 21) than the number 

 of cases of fuller regeneration (5 mesenteries). Toward incom- 

 plete bounding mesenteries of second or third grade, no reduced 

 numbers are found; and the same is true of regeneration follow- 

 ing divisions in exocoels. No reductions of the type here con- 

 sidered are recorded for regenerations after division in directive 

 endocoels. The instance of regeneration of but five complete 

 mesenteries under these conditions is no. 10b of table 3. This 

 specimen shows an incomplete new bounding mesentery in place 

 of the directive that would be expected to pair with the old 

 bounding directive. 



It may be well to call attention here, on the one hand, to the 

 rather fixed character of mesenteries and other structures added 

 in regeneration — a set of structures only slightly variable except 

 as modifications near the edges are necessary to enable this set 

 to fit into normal Hexactinian order with old parts adjacent to 

 the boundary — and, on the other hand, to the highly variable 

 result of such a regenerative process superimposed upon the 

 process of fission previously described. The extremely variable 

 number of complete mesenteries is thereby fully explained. 



From what has been said of the number of mesenteries found 

 in the new region, it is clear that the production of new mesen- 

 teries does not continue indefinitely. On the contrary, the 

 result of the process is strictly limited. In the following pages 

 it will be shown that the new mesenteries appear in a quite 

 definite order. 



Order of development of mesenteries 



For my study of the order of development of mesenteries in 

 regenerating regions, I have used such naturally divided speci- 

 mens represented in tables 3 to 6 as showed sufficiently early 

 stages of regeneration, and also some specimens that had been 



