in Lilium Mart agon : II. Spermatogenesis . 1 99 
hand, we again have evidence as to the seriation of the 
developmental stages from the position of the mother-cells in 
the loculus. Moreover, the smaller diameter of the nuclei is 
of advantage in one way ; for though an untouched nucleus 
is only to be met with in hand-sections, yet serial sections of 
15 ijl thickness include more than half the depth of favourably 
placed nuclei, and such preparations form a valuable link 
between those which show the complete nucleus, in which 
details are obscured by the thickness of the section, and the 
clear but fragmentary preparations of nuclei found in serial 
sections of thickness varying from 5 fi to 10 
As in the embryo-sac nucleus, longitudinal fission is found 
in parts of the spirem-ribbon before it has been divided into 
lengths by transverse division ( x and x Fig. 5) 1 . Articulation 
or beading of the spirem-ribbon precedes the formation of 
these loops. Short segments of the ribbon, including two 
or more pairs of granules, become slightly swollen, and the 
ribbon between such segments is rather narrower than in other 
places (Fig. 4 a). In the preparation from which Fig. 5 is 
drawn, indications are already seen of the approaching solu- 
tion of the nucleolus : the nuclear sap is coloured cloudily in 
places. Thus at x the space between the two rows of granules 
is slightly coloured as compared with the surrounding space, 
but this only occurs in the neighbourhood of the nucleolus. 
At x , for example, the space within the loop is quite clear. 
Details of the process of longitudinal fission are seen in such 
tangential sections of nuclei as that shown in Fig. 5#, which 
is from the same anther as the nucleus drawn in Fig. 4 a. 
The separation of the two rows of granules is nearly if not 
quite complete when the spirem-ribbon first shows indications 
of a division into lengths. Fig. 6, for example, is drawn from 
a nucleus near that from which Fig. 5 is taken ; at x a trans- 
verse division has just taken place, and when the section is 
narrowly examined, longitudinal fission can be made out 
in the ribbon wherever the nuclear sap is uncoloured. 
Solution of the nucleolus ( n ) has clearly begun, and the 
1 J. B. Farmer, Journal of the Royal Microscopical Soc., Oct. 1895, p. 502. 
