MR. E. S. GOODRICH ON SYLLIS VIVIPARA. 107 
ever that the yolk stored in the egg is not enough to provide 
nourishment sufficient for the growth of the embryos to the large 
size they attain in the body-cavity of the parent. Further food- 
material the young must derive from the mother during develop- 
ment; and, since there 13 no special connection of tissues between 
the two, this food would appear to be taken in either by mere 
diffusion through the general surface of the body, or by the 
mouth or anus. In some such way the embryos must obtain 
nourishment from the ccelomic fluid of the mother. 
The second question is less easy to answer. In the first place, 
since all the specimens I found contained ova, it is just possible 
that we have here a case of parthenogenetic development *. 
Should, however, spermatozoa actually at some time penetrate to 
the ceelom and fertilize the ova, it is by no means easy to see 
how such a process couid take place. The surface of the body is 
covered with a cuticle, and it is not likely that the spermatozoa 
reach the celom by entering the alimentary canal and boring 
their way through its wall. The only other means of access 
would appear to be the nephridia. Now these organs, so far as 
T have seen, do not undergo any of those changes in size and 
structure at maturity, such as are known to take place in many 
allied forms. The nephridia are delicate tubes, with a very 
narrow lumen, opening internally by a small nephrostome, and 
externally by a minute pore. The cilia of the nephrostome and 
of the canal act in such a way as to force substances down the 
canal to the exterior. That spermatozoa struggle up this narrow 
and difficult path is hard to believe. 
One possibility remains to be considered, though it must 
be admitted to be a somewhat improbable one—namely, that it is 
a case of self-impregnating hermaphroditism. A few viviparous 
hermaphrodite Polychztes are known, such as Neres diversicolor, 
Salmacina Dysteri, and Pomatoceros triqueter ; and it has been 
suggested by Schroder for the first species (7), and by de Saint- 
Joseph for the two last (6), that in these worms self-impreguation 
occurs. 
The evidence of hermaphroditism in Syllis vivipara is by ao 
means convincing. In sections of the hinder segments, whilst 
* MM. Mesnil and Caullery have already suggested that in one viviparous 
form of the Cirratulid Dodecaceria concharum the development of ihe ova is 
parthenogenetic (5). 
§* 
