ASEXUAL KEPRODUCTION IN PTYCHODERA CAPEXSI:^. 395 
process of asexual r(?prodiiction was continuous in the summer months, but 
not rapid. In the case ol: half-a-dozen P. proliferans, kept in a separate 
vessel, about four to six small fragments were given off every week. For a 
time these remained in close proximity to the parent, bnt many of them were 
subsequently found free on the suri'ace of the sand, and were readily moved 
about by any commotion of the water — probablj' a means of dispersal. 
Some of these fragments were isolated and examined from time to time. 
Thoy moved about freely, a proboscis and traces of a collar appeared, and 
for three weeks seemed to be developing into a form like their parent 
P. proliferans, but, about the end of this period, some of them were seen to- 
be developing hepatic caeca and genital wings, while still retaining the bright 
yellow colour and narrow intestine of the parent. Finally, after about a 
month, they had the general characteristics of P. capensis, with a few 
well-marked cseca and whitish caudal region, now containing a few grains of 
sand. The bright yellow colour had entirely disappeared even from the 
branchio-genital region, and they could not be distinguished from small 
P. capensis at a stage in which the gonads are not yet developed. 
Fig. 3. Fig. 4. 
Proliferated part 14 days after Proliferated part 25 days after 
separation. separation. 
In order to ascertain further details of the transformation of the one form 
into the other, various stages were examined by sectioning. Sections of the 
tail region of P. proliferans show that the lateral septa are well developed 
throughout its length, and are in intimate connection with the gonads, which 
contain eosinophil globules, and, in some cases, one or two ova. The same 
condition is^f' of course, found in the segments which break off from its 
posterior extremity, and this is maintained until the stage when the proboscis 
with notochord, &c., have appeared. 
Fifs. 5 and 6 show the conditions in a segment in which the proboscis is 
clearly • shown, but not the collar region, and the gonads occur, not as a 
continuous streak, but at intervals, so that in some sections' (fig. 5) the 
gonads still occupy a large part of the body cavity, but a few sections 
further on (fig. (>) they are almost entirely absent, though the lateral 
septa are still present. Sections of the same region at a later stage (about 
4 weeks) (fig. 7) show an entire absence of gonads and no trace of lateral 
septa. Fragments of food-material — diatoms, protozoa, and flocculent 
oro-anic matter — may now be seen in the intestine. 
Evidently, therefore, the two species are identical, P. proliferans being 
capable of giving rise asexually to P. capensis. The question then arises as 
29* 
