146 John H. Gerould, 



The conditions that are found in Sipunculus could have arisen 

 only from forms like Phaseolosoma and the mesotrochal Chaetopods 

 with large, but unmodified, prototroch cells. Adaptation of the 

 larva to active pelagic life, accompanied by loss of yolk, would 

 account for the modifications that have appeared in the development 

 of Sipunculus. 



Comparisons with Chaetopods show that striking resem- 

 blances exist between the mesotrochal trochophore of Amphitrite, of 

 Psygniohranchus, etc., and that of the Sipunculids. 



Evidences of metamerism appear in the development of Phaseo- 

 losoma, viz. paired lateral bristles (according to Selenka's account 

 of a form found at Villefranche) and transitoiy internal metamerism 

 in the rudiment of the nerve cord and mesoblastic bands in Ph. 

 gouldii, but it should be noted that the rudiment of the nerve cord 

 in Sipunculids is unpaired, unlike that of Echiurus and other Chae- 

 topods. 



Echiurids are probably degenerate Chaetopods, though there is 

 some evidence that they may be primitive. Sipunculids differ from 

 them in the unpaired rudiment of the nerve cord, in the character 

 of the retractor muscles, position of the anus, absence of anal 

 vesicles, and less evident and fewer transitory somites in the trocho- 

 phore. The Sipunculids show in their development no stronger 

 resemblances to the Echiurids than to other Chaetopods. 



Comparisons with Molluscs. The trochophores of Sipun- 

 culids resemble closely those of certain of the more primitive 

 Molluscs. The modifications of the prototroch in Sipunculus are 

 strikingly similar to modifications of the velum in Solenogastres and 

 in certain primitive lamellibranchs. The invaginations in the 

 posterior lip of the stomodaeum in Chiton, and other molluscs, cor- 

 respond in form and position to similar infoldings in the trocho- 

 phore of Sipunculus, viz. an anterior (glandular organ) and a posterior 

 (radular sac in Chiton, Schlundkopf in Sipunculus). 



Sipunculids are closely related to the more primitive molluscs, 

 though they resemble still more completely the Archiannelida 

 and the Chaetopods. 



Comparisons with Verm idea. Sipunculids are much less 

 closely connected with Phoronis, Bryozoa and the Brachiopods than 

 with Echiurids and other chaetopods, the development of these 

 vermideans being widely aberrant from that of Annelids and Molluscs. 



There is no embryological evidence at present in support of 



