136 G. HERBERT FOWLER. 



hooks.'"'' My specimens showed never less than three or more 

 than four terminal branches. The branching of the lateral 

 spines seemed always to be regularly dichotomous, and to end 

 in from eight to thirty-two fine " tuning-forks " (Dr. Borgert's 

 threads). Of the two whorls of hooks which he figures 1 have 

 once or twice caught a glimpse in making preparations, but 

 they have almost invariably been broken away in my speci- 

 mens. 



The internal spines of the region of the shell-mouth are 

 somewhat complex. In the anterior portion of the mouth, 

 all the way round, a ring of spines (sp. i) radiates inwards 

 below the edge of the lip across the circular opening of the 

 shell; these branch to form a fine meshwork supporting the 

 extra-capsular protoplasm and plucodium. Tlie posterior 

 portion of this ring passes over the lip to the outer edge, and 

 is directed outwards, the spines here being shorter and more 

 numerous (sp. ii) ; these also branch to form a meshwork 

 which lies on the Hoat, and, as far as I can ascertain, becomes 

 actually continuous with it, fastening float to shell. A few 

 spines, given off from the body of the shell itself (sp. iii) and 

 from the bases of the most posterior arms i-ight and left, are 

 similarly inserted on, and help to attach, the float. Spines 

 from the float itself also ramify into the extra-capsular region 

 (sp. iv). A meshwork of very hue character takes origin also 

 from the lowest or pitted zone of the shell-mouth, and ramifies 

 in and on the diaphragm ; a specially strong bundle of fine 

 spines belonging to this system, left after treatment with 

 strong warm acids, is generally to be found on the inner 

 anterior aspect of the mouth (sp. v). Presunuibly all the 

 meshworks of the plia3odial region arc continuous with one 

 another; but as they disappear entirely in cleaning the 

 skeleton of plueodium by any method that 1 have tried, they 

 can only be nuidc out by sections and teased pre})arations ; 

 the sections, at any rate, appear to point to continuity 

 throughout. 



The FLOAT is nearly spherical, but s(-)mewhat llattened or 

 dinted on the side which faces the pha?odial region ; it is 



