707 
Beer. — Studies in Spore Development. II. 
what smaller ones. The nuclear reticulum has a peripheral arrangement and 
the nucleolus (or nucleoli) appears to lie in a cavity. When such a nucleus is 
viewed in section one can occasionally find two chromatic aggregates or two 
filaments of the reticulum which lie parallel with one another. This, however, 
is merely due to the sides of the meshes of the network appearing to be closely 
approximated when seen in section. In a surface view of such a nucleus, 
this appearance of a parallelism of its parts usually vanishes altogether. 
Nuclei of Doronicum at this stage are shown in Fig. i, a and b. 
In slightly older nuclei, that is to say, in nuclei which are nearer to 
synapsis, the reticulum has become somewhat finer and smoother. The 
chromatic aggregates have become much less evident and the thickened 
bars or meshes of the network tend to become more even. The nucleolus 
still lies in a clear cavity. It will be seen from Fig. 2, which represents such 
a nucleus, that the filaments of the reticulum often have a dotted appearance. 
This is due to the points of crossing of the filaments or the nodes of the 
network being seen in optical section, and does not represent a granular 
arrangement of chromatin particles within the linin threadwork. At this 
stage the delicate meshwork shows no distinction between chromatin and 
linin. 
In Tragopogon pratensis the behaviour of the nucleus during pre- 
synapsis is very similar. Here also a rather coarse network tends to become 
smoother and more even before synapsis is entered upon. 
In Tragopogon the reticulum often appears to reach up to and to be 
attached to the nucleolus, so that this body is not in such cases surrounded 
by a clear space. Fig. 3 shows this relation between the nucleolus and the 
reticulum. Mottier (18) has described a somewhat similar appearance in 
some of the nuclei of Lilium Martagon at a corresponding stage. He 
mentions, however, that the chromatin lumps are collected round the 
nucleolus in Lilium , but I do not find this in Tragopogon. 
Calendida officinalis quite agrees with the Compositae which I have 
already described. A rather coarse network with chromatin aggregates 
which are irregular in size and form, and of no constant number, becomes 
more even and with less obvious chromatin lumps upon it before synapsis is 
reached. Fig. 4 shows these nuclei. Crepis virens i Crepis taraxacifolia , 
and Anthemis cotida are in no way different. 
In Calendida and in both species of Crepis the nuclear reticulum is 
peripherally arranged and a clear space lies round the nucleolus. 
The description which I have given above and the drawings which 
illustrate it will show the striking differences which exist between the 
appearance of my preparations of these stages and the figures and account 
given by Rosenberg and Lundegardh. 
For example, I have searched my preparations of Calendula repeatedly 
for a resting nucleus of a pollen mother-cell containing a number of regularly 
3 A 2 
