The Sexual Phases of Myzostoma. 
273 
growing oocyte and both bodies are invested with 3. a layer of 
fiattened peritoneal cells. The latter form a pedicel which at first 
anchors the egg and Zellenknopf to the wall of the ventral blood- 
vessel, but which fìnally ruptures and allows the egg to float ont 
into the body-cavity. The centrai celi degenerates and ultimately 
the whole Zellenknopf together with the investing peritoneal layer 
falls away from the ovum; at least Seengel records the fact (pag. 370), 
»dass man gelegentlich in der Leibeshöhlenflüssigkeit Zellenballen 
antrifft, welche die größte Ähnlichkeit mit einem Zellenknopfe haben ; 
ich zweifle nicht daran, dass es in der That von den Eiern abgelöste 
Zellenknöpfe sind«. 
In another Gephyrean, Thalassema^ if we may accept an Ob- 
servation of Semper quoted by Ludwig ('74), the conditions seem to 
be much simpler and more like those observed in Ophryotrocha ; the 
Zelleuknopf being reduced to a single cell. Leydig ('49) long ago de- 
scribed and figured a mass of cells attached to the egg of the Hiru- 
dinean Piscicola and evidently comparable to the Zellenknopf of 
Bonellia. Ludwig, too ('74, pag. 67 and Taf. 1 Figs. 9 and 10), called 
attention to similar accessory cells in Pontohdella and Branchellion. 
All the Chaetopods here considered agree in possessing Nähr- 
zellen which accompany the ovum into the body-cavity and only 
leave it, when it is nearly or quite full-gvown. In number and 
arrangement, however, these cells vary greatly in the different species, 
from the simple condition seen in Ophryotrocha and Thalassema to 
the elaborate Zellenknopf of Bonellia and the long cell-strings of 
Diopatra. Probably an investigation, undertaken with this particular 
purpose in view, would bring to light in many other Chaetopods 
still other peculiarities in the number and arrangement of the accessory 
cells. Be this as it may, however, the described forms differ from 
the accessory cells of Myzostoma in two respects: 1. in Myzostoma 
there is an accessory cell at either pole of the oocytes, 2. the ac- 
cessory cells of Myzostoma are soon worked into the cytoplasm of 
the growing oocyte and not ultimately cast off into the body-cavity 
after prolonged association with the ovum. These differences seem 
at first sight to be very important and seriously to invalidate the 
homology which I am maintaining, but the following considerations 
show that such is not really the case. In the first place I am not 
perfectly sure that all the oocytes of Myzostoma have two accessory 
cells; in several instances I have found only one, and, while I am 
inclined to believe that the other cell may be concealed behind the 
