1907.] 



NATURAL SCIENCES OF PHILADELPHIA. 



129 



layer, but saw only that i)orti()u of it which lios veiitrad of the ovary, 

 the dorsal portion beinii concealed by the close apposition of tho 

 ovary and the alimentary canal. l'\)r this reason he interprets it as 

 a portion of the splanchnopleure, api)arently assuming; with RepiacholT 

 (ISSG) the iiresence of a continuous i)eritoneal lininj;- throughout the 

 body cavity. The ovary would then, accordini;- to von Malsen, lie 

 outside of the ccrlom, between the alimentary canal and the peritoneum. 

 The presence of a definite peritoneal investment surrounding the 

 ovary of D. conklini is in accord with the observations of both Harmer 

 (1889a) and Schinikewitsch (1895). In the two forms investigated 

 by these authors — D. tcvmatas and J), vorticoidcs — the gonads consist 

 of long paired sacs, from the walls of which arise the sexual products. 

 These sacs, moreover, in D. vorticoidcs, according to the researches of 

 Schimkewitsch, arise from the paired mesoderm bands. There can, 



Fig. V. — Diagrams illustrating the reduction of the sex glands in DinopTiilus, 

 as represented by A, D. vorttcoides and tceniatus; B, D. gigas; C, D. conklini. 



therefore, be little doubt that the gonads with their enclosed cavities, 

 as^Meyer (1901) has also pointed out, are homologous with the peri- 

 toneum and coelom of the Annelida. 



It is evident, however, that, in the light of this interpretation, the 

 coelom of D. conklini is greatly reduced, as compared with those species 

 mentioned above. This reduction becomes more comprehensible 

 when the condition in D. gigas is taken into consideration. The sex 

 gland of this species is, according to Weldon (1886), Y-shaped, the 

 unpaired limb lying posterior, the two anterior limbs extendino- for- 



