Further Studies on the Ovogenesis of Sagitta. 247 
was prevented by some unusual conditions, and that the usual order 
of events so far as the distribution of the chromosomes is concerned 
went on, though for some unknown reason the polar body was not 
given off. No other peculiarity was observed. 
The question now arises whether the sections throw any light 
on the method of reduction in number of chromosomes in the young 
oocytes. This is one of the questions that I had not been able to 
answer in my earlier work, but Sagitta elegans proved to be more 
favorable material than Sagitta bipunctata. Among the young oocytes 
are some which contain the somatic number of chromosomes (18), 
some V-shaped (Fig. 18), others shortened and irregular in form 
(Fig. 19). A few of these showed a part of the chromosomes 
arranged in pairs (Fig. 20). A somewhat later stage shows the 
reduced number of chromosomes (9) elongated, and some have a 
longitudinal split (Fig. 21 and 22), which evidently corresponds to 
the space between two chromosomes of a pair in the earlier stage, 
and probably to the longitudinal division shown in Fig. 13 and to 
the wider split in the tetrad of the first polar spindle (Fig. 14 a). 
In Fig. 23 is shown an oocyte in which such pairs of chromosomes 
as appear in Fig. 21 are seen in cross-section. 
In the spermatogenesis of Sagitta we have the reduction effected 
by pairing of the chromosomes end to end, and the first maturation 
division separates the two chromosomes, while the second probably 
completes the longitudinal division that has been observed during 
the growth stage of the spermatocytes, between the two stages 
shown in fig. 28a and b, tab. 21 (1903) (Fig. 24a and b). 
In the ovogenesis the method of pairing is different and the 
order of the maturation divisions is reversed, the second division 
separating the components of the pairs. From this it would appear 
that it is the “conjugation” of the chromosomes and their longi- 
tudinal division and separation that are important, and not the 
method of union or the order of divisions. There is need, it seems 
to me, of going over as carefully as possible the ovogenesis of all 
forms which have been described as undergoing, during the process 
of maturation, a double longitudinal instead of a longitudinal and 
a transverse division, for evidence whether or not the two kinds of 
divisions are equivalent. 
Since in Sagitta we find two methods of “conjugation” of the 
chromosomes, — pairing end to end in the spermatocyte and longitudinal 
pairing in the oocyte; and the corresponding two types of maturation- 
