495 



has a very definite margin anteriorly (Fig. XIII, 1). In surface view it 

 shows delicate striations at right angles to this margin. As the margin 

 frequently shows distinct teeth the striations are doubtless due to a 

 corrugation of the surface. As the spinule is very minute, (usually from 

 10 to 30 \i long) it is difficult to properly interpret what is seen. The 

 next stage is shown in Fig. XIII, 2. The anterior margin projects 

 forming a lip, inside which is a groove. When the lip becomes more 

 prominent a definite pocket is formed inside it (Fig. XIII, 3, 4 and 5). 

 The margin of the lip may be truncate (2 and 4) or pointed (3 and 5). 

 It may be smooth (2 and 3) or dentate (4 and 5). In only one species 

 that I have examined is a definite spine formed (6). It is very evidently 

 an elongation of the median point of the lip. 



Fig. XIII. Siphonal spinules of Styelids. (X ca. 660.) Cnemidocarpa joannae; 2, Pe- 



lonaia corrugata; 3, Oonioearpa coriacea; 4, Katatropa yakutatensis; 5, Styela 



plicata; 6, Katatropa greeleyi. In each case, the figure on the left is a surface view 



and that on the right is an optical median section. 



The spinules are found in all the forms having the testes growing 

 away from the ovary. Of these Pelonaia shows the simplest condition. 

 Of the other forms (having the testes growing under the ovary) I have 

 been able to find definite spinules only in Cnemidocarpa joannae. Also 

 they were less perfectly developed than in any of those of the other 

 group. In this species the cell has originally only one nucleus but dur- 

 ing the formation of the spinule the nucleus divides and a multinucleate 

 condition is attained. This does not occur in the other group. 



Beyond corroborating the classification based upon other charac- 

 ters, the spinules cannot be readily used for defining the genera. 



The shape of the body and the character of the surface of the 



