The Development of Phascolosoma. 109 



their development, but only after contractions had begun. „Unter 

 diesen Umständen", he adds, „ist der Vorgang-, durch welchen sie 

 sich vom Hautmuskelblatt abheben, nicht genauer zu verfolgen, und 

 auch der Bau dieser Muskeln kann erst an der durchsichtigeren 

 Larve beobachtet werden". 



The coelomesoblast in Phascolosoma is still in a solid, undifferen- 

 tiated condition after the retractor muscles are formed between it 

 and the ectodermic wall of the trunk. At the beginning of metamor- 

 phosis, when the diameter of the trunk is rapidly increasing, and 

 the somatic layer of mesoderm is being carried outward with the 

 bulging of the body wall, the retractor muscles may readily pass 

 through this loose layer into the incipient coelom. 



Each retractor muscle retains its connections with the ectoderm 

 in front by a cluster of elongated sensorj"- cells. Thus the muscles 

 and the sense organs through wiiich they are stimulated arise from 

 common rudiments, which are in some respects similar to the Neuro- 

 muskelanlagen w^hich Kleinenbekg (1886) and E. Meyer (1901) 

 have described in Lopadorhijnclms. I shall describe the form and 

 distribution of these neuromuscular structures in connection with 

 the metamorphosis, when they reach their full development and 

 begin to be functionally active. 



Two pairs of accessory retractors are found in the elongated 

 trochophore, in connection respectively with the chief dorsal and 

 ventral retractors (Fig. 88). They each consist of groups of two or three 

 cells of the zone immediately in front of the cells of the postoral circlet. 

 They become elongated, and from the deeper extremity of each a 

 long muscle process grows backward between the mesoderm bands 

 and the ectodermal wall of the body, into which, near the posterior 

 end of the trunk, the branching tips of the process are inserted 

 (Fig. 88). The ventral accessory retractors have branching processes 

 in front, wiiich are inserted into the sides of the stomodaeum, and 

 simple fibrous processes behind, which are attached to the walls of 

 the posterior end of the trunk, near the termination of the chief 

 ventral retractors. The former serve to draw backward slightly 

 the wall of the stomodaeum; and the earliest muscular movement 

 which I have observed in the trochophore was a spontaneous twitching 

 of the wall of the stomodaeum, which must have been due to the 

 action of these muscles. 



Circular muscle fibres (Fig. 85) are early developed in the trunk 

 from ectomesoblast cells which, lying immediately beneath the ectoderm, 



