1907.] NATURAL SCIENCES OF PHILADELPHIA. Ill 



may naturally he said of Dinophilus. Moreover, two other special 

 features, nanu'ly. the separation of the lateral halves of the ven- 

 tral nerve corel and the subdivision of the transverse commissures, 

 it shares with many Hermellidce and Serpulid(e (Meyer, 1886). Both 

 the epithelial position of the nervous system and the separation of the 

 nerve cords are found in Nerilla (Pereyaslawzewa, 1896), w^hich is 

 unquestionably a polychsetous annelid, while with the nervous system 

 of i'Elosoma (Brace, 1901), the most primitive member of the 

 Oligocha^ta, there is the closest resemblance. For example, A. tene- 

 brarum, described by Miss Brace, possesses not only a brain closely 

 soldered to the hypodermis, but also widely separated ventral nerve 

 cords, metamerically arranged groups of ganglion cells, and two or 

 more transverse commissures for each segment, all completely 

 embedded in the ventral hypodermis. 



This widespread ectodermal position of the nervous system is of 

 course readily explained as a retention on the part of these forms of an 

 embryonic condition — a permanently arrested stage of development — 

 apparently correlated with their small size and the resulting simplifica- 

 tion of their economy. That the same explanation will apply to 

 Dinophilus is sufficiently obvious; the more so in the light of the 

 researches of Schimkewitsch (1895) on the embryology of this form, as 

 well as the evolution of the nervous system, as seen in the difTering 

 stages of development shown by the various members of the group, 

 and described above. 



There remain three special features of the nervous system of 

 D. conklini. The first of these is the presence of a preoral transverse 

 commissure. This structure may be accounted for in one of several 

 ways. In the first place it might be supposed to represent a com- 

 missure originally postoral, and belonging to the trunk, but which 

 has become preoral in position by the shifting of the mouth in a 

 caudal direction. This would be a perfectly allowable assumption, 

 supported by analogy with related forms, since in both the flat worms 

 and annelids the position of the mouth is by no means a fixed one, 

 but quite subject to shifting. However, although the posterior limb of 

 the triradiate mouth does extend caudad into the trunk region, the 

 evidence of any considerable movement caudad on the part of the 

 mouth as a whole is too slight to lend this explanation of the occur- 

 rence of a preoral transverse commissure any considerable degree of 

 probability. 



In the next place this commissure may be regarded as having arisen 

 entirely de novo in response to the peculiar physiological needs of the 



