382 



TORREY 



Fig. 13. Edxvhrdsiella sipun. 

 culoides. Diagram of arrange- 

 ment of mesenteries and tentacles 

 (indicated by black spots). 



region (fig. 14) according to the line x-x (Fig. 14) shows that 

 although the mesenteries are minute, they are furnished with a central 



band of mesogloea, upon which is 

 spread a thin and rather inconspic- 

 uous, yet vuimistakable layer of longi- 

 tudinal muscles. 



The arrangement of the rudimentary 

 mesenteries and their accurate correla- 

 tion with the tentacles, shown in fig. 

 12, is not the invariable rule. Another 

 individual with thirty-two mesenteries 

 had but twenty-five tentacles. The 

 mesenteries were distributed according 

 to the formula : I, 3, II, 5, III, 5, IV, 

 V, 4, VI, 5, VII, 2, VIII, in which 

 the roman numerals represent the pri- 

 maries and the arable the number of 

 rudimentaries between every two pri- 

 maries. Not only are the rudimentaries not distributed according to 

 the numerical plan in fig. 13, but they are not even bilaterally arranged. 



If we examine the three 

 mesenterial types which 

 have been described for E. 

 beautefnpsi^ E. adene?zsis 

 and E. sipunculoides re- 

 spectively, it is at once evi- 

 dent that the plan exempli- 

 fied in E. bea7itempsi is the 

 only one that fulfils the re- 

 quirements of an ancestral 

 form. We may conceive the 

 transformation from the oc- 

 tamerous to the hexamerous 

 condition to have been ac- 

 complished by the growth of 

 the mesenteries in the lateral 

 and ventro-lateral spaces, 



the mesenteries in the dorso-lateral spaces forming the first two pairs 

 of the second cycle. 



The rudimentary mesenteries in E. adenensis arise apparently in 

 pairs which in arrangement suggest the second cycle in Halcaryipa. 



^ 2 2 



Fig. 14. Ediuardsiella sipunculoides. 

 Cross section through region of rudimentary 

 mesenteries, at level x-x in small figure. 



