A Contribut. to the Embiyol., Life-histoiy, and Classificat. of the Dicyemids. 77 



always mudi sraaller thau the tliree dorsal (one median and two 

 lateral) and two ventral metapolars (p. 9 aud 26 — 27). Calotte per- 

 manently orthotropal. 



The parapolars (p. 8) are two and lateral, in both genera. The 

 caudal cells are always two, and dorso- ventral, at least in the youngest 

 iudividuals. 



3. The dorsal and ventral sides of the calotte are not distingiiish- 

 able by any uniform diiferenee in the size of the polar cells. In Dicye- 

 mennea, the three dorsal metapolars serve as a means of orientation. 

 In Dicyema , the plagiotropal calottes face the ventral side ; the sides 

 of the orthotropal calottes are determined by the relative position of the 

 paiapolars and the adjoining ectodermal cells (p. 10). 



4. The obliquity of the octamerous calotte and the inequality of its 

 cells are extremely variable among individuals of the same species ; and 

 the range of variability can seldom, if ever, be ascertained from exam- 

 ples taken from a single Cephalopod (p. 12, 17, 21). It is uecessary 

 therefore to discriminate carefully between individuai peculiarities and 

 specific characters (p. 2). 



The ectoderm furnishes ali the diagnostic characters : the whole 

 number of ectodermal cells , the length of adult individuals, the size of 

 the polar cells both relative and absolute , their shape, position, and 

 axial relations, the form of the parapolar and caudal cells, etc., supply 

 the means of distinguishing species. 



5. When two species occur in the same Cephalopod, they may oc- 

 cupy separate renai Chambers and thus be completely isolated; or they 

 may be cohabitants of the same Chambers. In the latter case they are 

 seldom promiscuously associated, but distributed in colonies (p. 14, 28). 



If only one species is present, a colonial distribution is often observ- 

 able, some renai lobes being occupied more or less exclusivel}^ by ex- 

 tremely long individuals, others by shorter individuals. Sometimes the 

 entire renai organ , with the exception of one or two lobes will be free 

 from the parasite. We may infer from these facts that they are not 

 introduced into the host in large numbers at any one time. 



6. The Dicyemids may be divided, according to the share they take 

 in the work of reproduction, mio monogenie and diphy (/ente (ordiplogenic) 

 individuals ip. 55). The first class (primary Nematogeus) (p. 36), pro- 

 duce only vermiform embryos; the second produce first iufusoriform em- 

 bryos, then vermiform embryos (secondary Nematogens) (p. 41) . Rhomb- 

 ogen and secondary Nematogen denote two difterent phases in the lite 

 of the same individuai. It is stili doubtful whether the monogenie and 



