122 University of California Publications in Zoology. [Vol.6 



orders, they are grouped at the oral end of each mesenterial 

 filament. They are scattered along- almost the full Length of the 

 mesenteries of the first and second orders, beginning with a 

 compacl bunch on each mesentery just below the aboral ends of 



the coiled mesenterial filaments of the shorter mesenteries. 



There is nothing essentially different from this in the distribu- 

 tion of these structures in C. membranaa us. On the mesenteries 

 of the third ami fourth cycles, however, are associated with them 

 other structures of a type hitherto undescribed for any adult 

 form, namely, the cnidorages described by van Beneden tor 



Fig. 3. — C. benedeni. Process of mesenterial filament loaded with 

 cnidorages. 



three pelagic larval forms. A cnidorage is a follicle packed with 

 nettle cells, or cnidoblasts, that may be of several sorts, the largest 

 and by far the most important equalling in length the diameter 

 of the follicle and far surpassing all the others in bulk. In 

 C. benedeni, the cnidorages are very numerous on the bunched 

 filaments, as shown, for a single filament, by tig. 3; they are 

 much less numerous on the scattered filaments. In the larval 

 forms examined by van Beneden, the filament does not appear to 

 have been prominent as such, as in the present species. The 

 cnidorages were grouped together, however, like grapes in a 

 bunch with little or no stalk. Such aggregates van Beneden 

 called botrucnides, pointing out that they occupied the position 

 of acontia and probably possessed the same morphological value. 1 "' 



io"J'ai domic le aom de Botrucnides a des formations tres curieuses qui 

 dans les u'cnrr*. Cerianthula, Hensenanthula and Calpanthula, occupent la 

 place ilcs aconties et posBedent probablement la meme valeur morpho- 

 logique." (p. 32). 



