Trow . — On Fertilization in the Saprolegnieae. 563 
seem to be still recognizable, being grouped apparently in fours (Fig. 33). 
This reminds one of the statements of Berlese (’ 97 ) as to the number of 
chromosomes counted by him in the oospores of the Peronosporeae under 
similar conditions — statements which certainly aroused considerable scepti- 
cism at the time they were published. 
The further ripening of the oospores and their germination were not 
investigated in this species. The observations recorded suffice to prove 
once more the familiar succession in the number of the nuclei, the formula 
1, 2, 1 being again demonstrated. 
The number of chromosomes. As already indicated above, a study of 
the first mitosis in the oogonium led to the conclusion that the number of 
chromosomes in the nuclei was certainly more than four, and probably 
eight. The more difficult study of the second mitosis makes it pretty clear 
that the number of chromosomes is reduced therein to four. At metaphase 
in side views of the spindle we see clearly two or three chromosomes — 
numbers which would naturally occur if there were really four chromosomes 
in the nuclear-plate (Fig. 18 k and 20 a). It is not without significance, too, 
that four chromosomes can sometimes be seen in each of the conjugating 
gameto-nuclei (Fig. 32). Stress need not be laid on the actual numbers, as 
there are several sources of error in such determinations, but the reduction 
is definite enough. Such a division is to be regarded as a reducing division ; 
a reduction in the number of chromosomes takes place in gameto-genesis 
in this plant as in most animals, and not in sporogenesis as in most plants. 
A detailed discussion of the physiological significance of such reductions 
may well be deferred until more information has been gathered as to their 
distribution amongst plants, and especially amongst the sexual and apoga- 
mous members of the Saprolegnieae. The view expressed by me in ’99 as 
to the significance of these divisions during gameto-genesis must, for the 
present, suffice. 
Anomalies. In the discussion of cytological problems a consideration 
of typical forms is of much more importance than that of anomalous ones. 
The conclusions reached by the modern cytological methods become quite 
untrustworthy in the presence of many anomalies, for the simple reason that 
anomalous and typical forms get grouped together to form an unreal 
series — a sequence is evolved which is purely imaginary. One must con- 
stantly bear in mind that the series of stages in development described in 
such a paper as this are not consecutive stages in the^ development of some 
one individual organ, but rather stages arranged consecutively with more 
or less success, according to the calibre of the observer, representing a 
series of such organs. With this difficulty always in view, great care has 
been taken in the selection and cultivation of the material employed in my 
researches on the Saprolegnieae. In certain preliminary studies, on material 
afterwards rejected, anomalies were only too abundant. 
