ANEMONES 387 



free edge of the oesophagus. Those on the second cycle follow the 

 borders of their mesenteries farther dorsalward than those on the first 

 cycle, in which the mesenteries are wider. The upper ends of the 

 thick lines in the diagram should be taken as the summits of these 

 archings of the filaments and not as their points of attachment to the 

 oesophagus. 



There are several points of interest connected with the mesenteries 

 of the first cycle. First, there are two fairs of directives although 

 there is but one siphonoglyph. These directives are associated with 

 two slight downward prolongations of the oesophagus, and their 

 mesenterial filaments do not arch orally, but proceed immediately 

 downward along the edges of their mesenteries. The directives thus 

 appear to be the widest and inost important of all the mesenteries. 

 This importance is not associated with any structural peculiarity of the 

 mesenterial filaments, so far as I could see. 



Further, the directives are the only pairs of the first cycle whose 

 members are equivalent in width, strength of retractors, length of 

 mesenterial filaments and size of gonads. The other pairs are arranged 

 bilaterally on each side of them. The member of each pair with the 

 smaller gonad, shorter mesenterial filament, etc., is nearer the pair of 

 directives which is associated with the siphonoglyph. Since there are 

 two pairs of directives, this bilaterality is clearly correlated with the 

 single siphonoglyph. Such a conclusion is interesting in connection 

 with the tendency of the non-directives in the monoglyphic individuals 

 of Metridium to mass themselves on the side of the oesophagus oppo- 

 site the siphonoglyph and a pair of directives. There it is impossible 

 to tell whether the behavior of the mesenteries is correlated with the 

 siphonoglyph or the directives. The condition in Harenactis elimi- 

 nates this difficulty and makes it probable that the siphonoglyph is of 

 more importance than the directives in determining mesenterial arrange- 

 ment. While we are ignorant of the real nature of the correlation, it 

 is probable that this is to be sought first among the causes of the vari- 

 ation in the number of siphonoglyphs, regardless of the directives. 



Family ACTINIID.E Gosse. 

 Actiniidai adhering to foreign bodies by a flat contractile base. 

 Column usually smooth, occasionally verrucose towards the upper part, 

 without cinclides. Margin frequently provided with acrorhagi, but 

 may be smooth. Tentacles numerous, usually long. Sphincter muscle 

 endodermal, diffuse, usually feebly developed. Perfect mesenteries 

 usually numerous, and all may be fertile, but sometimes the first cycle, 

 and more generally the directives are sterile. No acontia. 



