fifth \\.is concerned, bul in 1 1 1 « - youngest individual observed by Boveri the lust four couples 

 already formed. I «ras fortunate enough to obtain a still yo larva of . /. brp.chiola.ta 



in which but six mesenteries were formed, and by a comparison of tliis with older forms it was 

 determine, with ;i close approximation to certainty, the order of succession of the 

 couples, and this proved to 1"- identical with that which earlier observations had 

 shown to occur in the Hexactinians, and quite different froifl that of the later formed couples. 

 it appears, then, that there can be recognized in the Ceriantharia two sets of mesenteries, 

 the protocnemes, consisting of the first four couples, in which the order of succession is 

 similar to that occurring in the majority of Hexactinians, and a second, the deuterocm 

 consisting ol all the remaining mesenteries, which appear successively in couples in the di 

 intermesenterial chamber. This distinction was clearly recognized by both Boveri and im 

 and was represented by diagrams in our respective papers. As tliis matter is of some importance 



in connection with the proper understanding of adult Ceriantharia 

 I venture to append a diagram illustrating the arrangement and 

 i irder of succession of the two sets i rexl figure l), the protocnemes 

 being represented in red and designated by roman numerals, while 

 the second set, the deuterocnemes, are represented in black 

 and designated by arabic numerals. 



1 rom this diagram it will be seen that the order of 

 cession of the protocnemes, starting from the midventral line, is 

 III — I — II — IV. and the fust couple of deuterocnemes makes its 

 appearence in the intermesenterial chamber bounded by the two 

 i. ig . i. protocnemes belonging to the fourth couple. Comparing this 



ing the probable succession sequence w i tn that indicated bv BoVERl in his figures, a discrepancv 



the mesenteries in the Ceriantharia. 



i iv, protocnen* is observable in that those protocnemic couples which I have 



arded as being the third and fourth are taken by Boveri to 

 be the fourth and third respectively ; at least they are designated in his figures by d and c. 

 discrepancy, however, is an unimportant one, since Boveri had no opportunity for deter- 

 mining the actual sequence of these mesenteries, nor, indeed, does he discuss thcir probable 

 sequence. llis designation of them may be regarded as a matter of convenience rather than 

 as indication of their possible ol appearance. 



The sequence which I have found for the protocnemes is of great importance, for it is 

 identical with that characteristic of the majoritj of the Hexactiniae and, in all probability, with 

 that of the Zoantheae; and it points back to an ancestral stage with eight mesenteries, com- 

 mon to all these groups, and from which the ( erianthea have branched ofH in one direction 

 and the Hexactiniae and Zoantheae in another, as 1 have already maintained in previous papers. 

 It is this ancestral character which gives to the protocnemes their great importance and which 

 mak' recognition of protocnemes and deuterocnemes fundamental for the proper \m^\- 



; the adult Actiniaria. 



thors, v.w Beneden and Vanhöffen, however, reached conclusions regarding the 

 protocnemes of Arachnactis more or less at variance with that described 



