553 
Trow . — On Fertilization in the Saprolegnieae. 
of this rather difficult, and in the present connexion somewhat irrelevant, 
problem. A comparative critical study of nucleoli is obviously a desidera- 
tum. The recent work of Wager (’04) and Williams (’04) make it clear 
that the current conceptions of the structure of the nucleus and of its 
behaviour during mitosis will have to be modified in several respects. 
My observations, so far as they concern other features of the karyology, 
agree with those of Davis in the following features: — (i) the intranuclear 
character of the spindle, (%) the persistence of the nucleolus (as a much 
smaller body than the original central mass) up to the metaphase stage in 
mitosis, and ( 3 ) the absence of centrosomes and astrospheres. It is clear 
to me, however, that the number of chromosomes in the nuclei of the species 
is certainly greater than four and probably eight, and that the nuclear 
figures are not nearly so regular as those figured by Davis. Moreover, the 
divisions of the nuclei are not so exactly synchronous as in the cases 
figured by him. The arrangement of the daughter-chromosomes as in 
Fig. 1 6 a suggests that a second mitosis is to take place, at any rate in 
some cases. The conditions seem to be more complex in this species than 
in Davis’s apandrous .S', mixta. 
Whilst the first mitosis is in progress the layer of protoplasm lining 
the oogonium gets thinner and thinner and it attains its minimum thick- 
ness before the second mitosis is initiated. It is at this stage that the 
real cytological difficulties have to be faced. It is noteworthy that up to 
this point in the development of the oogonia, AcMya De Baryana does not 
differ in any fundamental feature of its cytology from the apandrous species 
investigated by Davis. Such slight differences as are noticeable are pro- 
bably traceable either to specific variations in the plants themselves or to 
differences in the methods employed in their examination. 
The second mitosis. Davis recognized the difficulties which naturally 
arise at this stage, but does not appear to have succeeded in solving the 
main problem. He seems to have overlooked the fact that the stage in the 
development of the oogonium immediately preceding that of * balling’ is 
readily distinguished in sections by the thinness of the parietal layer of 
protoplasm. His figures appear to be founded on the examination of 
tangential sections alone. Making all allowance for possible errors of 
interpretation, I am unable to reconcile his views with the observations 
made by me in Achlya De Baryana. Future investigations must decide 
to what extent these differences are founded on errors of observation 
and judgement, or on differences of fact. 
When the anaphases of the first mitosis are over a single row of nuclei 
may be observed in the thin parietal layer of protoplasm (Fig. 17 ). The 
nuclei are not so numerous as they ought to be, and as some of them — 
marked k , /, m — stain so badly as to leave little doubt that they are under- 
going dissolution we may infer that the absorption of the superfluous 
