ANEMONES 377 



Edwardsia type, the sixteen tentacles are arranged alternately in two 

 cycles, the primary eight representing the outer, the other the inner 

 cycle (fig. 5). So far as numerical grouping goes, the same plan ob- 

 tains in an undescribed Edwardsia from San Pedro, Cal. ; but I have 

 been unable to detect any alternation of the tentacles at their bases, 

 though in life eight bend inward and upward, alternating with eight 

 which bend outward. Yet there is one interesting difference. In the 

 San Pedro species and, I infer, in E. beauiempsi^ the two groups of 

 three tentacles open into the dorso-lateral primary enteroccels. In 

 Carlgren's diagram these groups open into the ventro-lateral primary 

 enteroccels. It is difficult to derive the hexamerous type from such a 

 form. This leads me to suppose that Carlgren inadvertently figured 

 the longitudinal muscle bands on the wrong sides of the mesenteries. 

 Faurot, in 1S95, seems to have found a like arrangement in E. beau- 

 tempsi^ to judge from the arrangement of secondary mesenteries which 

 he has figured.' For the latter agrees with the plan of the mesenteries 

 in the San Pedro Edzvardsia (fig. 6), in which one tentacle communi- 

 cates with each intermesenterial space. 



According to Carlgren's second or Edivardslella type (fig. 7), the 

 eight primary tentacles form the innermost cycle, one in each primary 

 enteroccel. There are twelve tentacles 

 in the second cycle, two in each pri- 

 mary enterocoel save in those bounded 

 by the directives. The third cycle is 

 external to the second. This type will 

 not include E. sipiinculoides^ in which 

 species there are two and only two 

 distinct cycles, regardless of the number 

 of tentacles ; and the members of each 

 cycle, at least in the cases where the 

 number of tentacles equals the number 



of mesenteries, alternate with each ^^^ „ ^T^ T- „ 



' _ tiG. 7. inhvarasteUa type 



other. Consequent!}', if there are two (Carlcrren). 



cycles in the typical Edwardsia^ there 



would seem to be no essential difference between the two genera in 

 the arrangement of the tentacles. I shall discuss the details for E. 

 sipunculoides below (pp. 379-3S3). 



For a time I deemed the distribution of the capsules of nematocvsts 

 to be of generic value, since in E. sipunculoides these capsules are 

 scattered irregularly over the column, while in the typical Edwardsias 



1 Arch. Zool. Exp. (3), in, fig. 7, p. 113. 



