174 



contained both ve imi form and infusoriform embryos in pro- 

 cess of development. This, of course disposes of van Be ne den's 

 view of dimorphic females, and indicates a transition between the two 

 reproductive phases. 



That the sequence is from the nematogenic to the rhombogenic 

 phase and not vice versa is also indicated by the detailed study of 

 the contents of the axial cell in the transitional stages. Only a portion 

 of the germ-cells gives rise to vermiform young in the nematogenic 

 mother, so a stage is reached in which the axial cell is filled with ver- 

 miform embryos plus a number of isolated germ-cells. This stage 

 is especially common in Octopus 4 — 5 cm long. I b elieve that many 

 of these germ-cells then leave the parent, for somewhat 

 later one finds the diapolar ectoderm cells each containing 

 as many as four or five nuclei. These resemble germ-cell nuclei 

 and can only come from the axial cell since the ectoderm nuclei are 

 never seen dividing even by amitosis. As these nuclei finally disappear 

 they must be either expelled into the renal liquid or absorbed by the 

 protoplasm of the diapolar cells. The germ-cells in the protoplasm of 

 the axial cell, thus greetly reduced in number, come to be separated 

 from one another by long intervals. They then undergo cleavage to 

 form the infusorigens , of which there may be as many as five or six in 

 one Dicyema^ and these in turn produce the infusoriform young. As 

 a rule all the vermiform embryos have left the parent before the in- 

 fusorigens appear, but as in the ten or twelve cases above mentioned, 

 the development of the two kinds of offspring may overlap, unmista- 

 kable infusoriform embryos coexisting with vermiform embryos in the 

 axial cell of the same Dicyema. 



2) As Balfour (1881) has said, the taxonomic position of the Di- 

 cyemidae must depend on the interpretation of the infusoriform embryo. 

 Properly speaking this is not an embryo at all, but must be either a 

 larval or adult form since it escapes from its mother and is able to move 

 about for some time in the renal secretion or in the sea-water. I agree 

 withvan'Beneden(1882)in interpreting this singular or- 

 ganism as the male Dicyemid. That there are strong grounds for 

 this view is evident when we compare the infusoriform with what is 

 certainly the male of the Orthonectidae, the nearest allies of the Di- 

 cyemidae. The male Orthonectid [Hhopalurcù] combines characters of 

 the vermiform, i. e. of the female, with those of the infusoriform Di- 

 cyemid. It is long, fusiform and radially symmetrical like the vermi- 

 form, but, like the infusoriform, contains a central body full off gra- 

 nules. Its ectoderm, moreover, is composed partly of smooth, eciliate 

 cells, each containing a refractive body, and partly of powerfully ciliated 



