STUDIES ON THE PEOTOOHORDATA. 327 



posterior walls of the stigmata, and, growing across the aper- 

 ture, fuse with the opposite wall. This process of subdivision, 

 which was correctly described by van Beneden and Julin (6), 

 is then continued, and we arrive at the stage with six trans- 

 verse rows of secondary stigmata derived by repeated subdivi- 

 sion from the six primary stigmata (cf. fig. 17). In the next 

 stage occurs the change in the direction of the axes of the 

 stigmata, which was first observed by Krohn. Hitherto 

 transverse, their long axes now become vertical or longitu- 

 dinal. A single row of stigmata in process of changing their 

 axes is shown in PI. XXXI, fig. 18. From this figure it can 

 be seen that the individual stigmata undergo a bow-like curva- 

 ture, and then divide in the hollow of the bow, the result 

 being that the new stigmata thus produced have approximately 

 a vertical instead of a transverse long axis. 



The next question relates to the way in which the numerous 

 transverse rows of stigmata which compose the adult branchial 

 system are produced. Van Beneden and Julin (6) appear to 

 think that, after the formation of the first six rows of secon- 

 dary stigmata, the stigmata which are formed later arise by 

 fresh perforations. What they actually say is as follows : — 

 " Ces premiers stigmates, au nombre de six, se subdivisent par 

 une sorte d'etranglement en stigmates secondaires. Six ran- 

 gees de stigmates secondaires naissent ainsi des six stigmates 

 primaires. A une periode plus avancee du developpement le 

 processus genetique primitif [this means the method of inde- 

 pendent perforation], reparait pour donner naissance a de 

 nouvelles fentes stigmatiques interposees entre les stigmates 

 secondaires formes par etranglement." 



In Ciona this is certainly not the case, and, judging from 

 the closely parallel development of the members of the 

 Phallusia group, it is difiicult to believe that the condition of 

 things can be entirely difi^erent. In fact, the new trans- 

 verse rows of stigmata arise by division of the pre- 

 existing transverse rows. 



This is easy to establish by an examination of any young 

 specimen of Ciona, as well as the fact that the very small stig- 



