706 Beer, — Studies in Spore Development . II. 
My own observations upon the pollen development of the Compositae 
were undertaken with the primary object of studying the structure and 
development of the membrane of these bodies. 
I saw no reason to doubt the correctness of the results reached by the 
Swedish investigators in their study of the reduction divisions, and I, there- 
fore, turned to my preparations of the dividing pollen mother-cells with the 
full expectation of finding that these would confirm the previous observa- 
tions. I soon found, however, that my sections presented so many and such 
notable points of difference from what I had anticipated from Rosenberg’s 
descriptions, that it was no longer possible to dismiss the nuclear phenomena 
of these plants in a few words, but that they would require a very careful 
re-examination. 
I have, therefore, decided to divide my account of the pollen develop- 
ment of the Compositae into two parts, and to deal in the present communi- 
cation with the nuclear divisions (both meiotic and somatic) alone, whilst 
I reserve the detailed description of the pollen-wall for a separate part of 
these ‘ Studies ’. 
In dealing with the meiotic phase I have examined Tragopogon praten- 
sis , Matricaria Chamomilla , and Crepis taraxacifolia , throughout the different 
stages. In addition to these, I have also carefully examined the earlier 
prophases of the heterotype division in Dorojiicum plantagineum^ Calendida 
officinalis , and Anthemis Cotida . The later stages in the telophase of both 
the heterotype and the homotype division were also studied in the case of 
Crepis virens. 
For comparison with the meiotic phase I have also made a careful study 
of the somatic divisions in Crepis virens } 
Presynapsis. 
In Doronicum plantagineum , as the nucleus passes over from the telo- 
phase of the last premeiotic division to the ‘resting’ condition* we find that 
the gradual vacuolation of the chromosomes and the dispersal of their 
substance lead to the formation of a coarse reticulum. At one spot this 
reticulum consists of only delicate fibres, at another the strands of the net- 
work are coarse and thick, and the chromatic material may even be 
aggregated into irregular clumps. There can be little doubt that these 
thickenings upon the reticulum and the irregular clumps of material are 
derived from the incomplete dispersal of the chromosome bodies of the last 
somatic division. There is, however* no constancy in their number, size, 
form, or arrangement, and they cannot be regarded as prochromosomes in 
the sense in which this word is used by cytologists (PI. LXVI, Fig. i). The 
nucleus at this time contains either one large nucleolus or two to three some- 
1 The material was fixed, with all precautions, in the stronger Flemming’s solution, and in an 
alcohol and acetic acid mixture. 
