ASEXITAL KEl'KOUUCTION IN PTYCHODEKA CAPENSIS. 'i07 
contained a few grains of sand. At a later date hepatic cajca appeared 
about the middle of the body, and subsequently all were transformed into 
typical P. capensls, with fairly well-developed gonads, in which, however, 
there were no ova. Thus both the parent and its proliferated parts ultimately 
assumed the form of P. capensis. 
Ptyclwdera proliferans would seem, however^ to be capable of reproducing 
sexually, as the gonads contain well-developed ova. The facts, however, that 
(1) ova ma}" occur in the proliferated part, and be used up like the granules 
in the growth of the body, and (2) that the ova ultimately disappear in the 
reduced adult, seem to indicate that there is no sexual reproduction in this 
form. There can be little doubt that P. capensis, into which both buds and 
parent are transformed, reproduces sexually, so that the whole piocess looks 
like a modified form of alternation of generations. There is evidence, how- 
ever, which seems to indicate that P. proliferans arises from P. capensis by a 
division of the latter in front of or at the hepatic region, and subsequent 
prolongation and proliferation of the genital region, followed finally by 
regeneration of the lost hepatic and ciiudal region. 
The nutritive eosinophil globules or granules evidently play an important 
part in the life-history of the animal, and their relation to the sexual elements 
may therefore be further enquired into. 
They have been noted in many other Enteropneusta, but their origin and 
function remain in doubt. Spengel found that, in P. minuta, they were con- 
tained in cells in which no nuclei nere seen. In P. flava Willey found no 
normal nuclei, and believed that the nuclei of the cells undergo a process 
of degeneration analogous to fatty degeneration. They were found to be 
present when the germ-cells were fully developed, but both authors found 
that they disappeared at the period of complete sexual maturity in the 
species they examined. As to their functions, ISpengel could come to no 
definite conclusion ; Willey suggested that they are partly for the nutrition 
of the growing germ-cells, but principally for providing an albuminous 
covering to protect the germ-cells during maturation. 
In P. capensis they are absent only in the young stages. They are present 
from the time the germ-cells begin to develop and increase in numbers with 
them, and are present when the animals are fully developed. They do not 
disappear at complete sexual maturity, though the sexual elements then pre- 
ponderate, and they subsequently (towards the end of the winter, when 
exceptionally large specimens were found) appear to increase in number, 
their bulk greatly exceeding that of the remaining germ-cells. 
In P. ptroliferans stage they constitute the main mass of the gonads. 
These are greatly attenuated posteriorly and occur, as already stated, as two 
narrow streaks of a bright yellow colour in the elongated posterior region of 
the body, which is rounded and without pleural ridges in the living animal. 
The relative abundance of ova and globules was ascertained by the methods 
