470 Far7ner . — On Spore-Formation and 
protoplasmic structures at the critical period of spore-forma- 
tion are unfortunately exceedingly difficult to fix properly, 
although this observation does not apply to the nuclei of any 
of the other cells, whether of gametophyte or sporophyte. 
For staining, I chiefly used Heidenhain’s iron-haematoxylin, 
followed by a- ground stain of Orange G. In this way it was 
possible to trace the developmental history of the chromo- 
somes, though the results obtained were checked by the use 
of many other dyes, used both singly and in combination. 
In all cases the microtome was used for cutting the sporogonia, 
which had been previously embedded in paraffin, but observa- 
tions were also made on uninjured spore-mother-celis which 
had been teased out of the sporogonium. 
Fossombronia. 
I examined three species of this genus, F. Dumortieri , F. 
longiseta 1 , and F. angidosa . I found the first-named species 
on the whole the best for my purpose, as I had a large 
quantity kindly sent me by Mrs. Tindall ; nuclear divisions 
were not so frequent in the material of the other two species. 
The archesporial cells are fairly large, and contain very 
prominent nuclei. The nuclei seldom include more than one 
nucleolus, which is of a relatively large size ; the linin consists 
of a freely anastomosing fine threadwork, which is very deli- 
cate in the resting nucleus. The anastomosing spots, and also 
intercrossing threads, give to the nucleus a granular appearance 
at first sight. The nuclear wall is very obvious. The proto- 
plasm outside it is of a granular texture, and in the resting 
cells I was unable to determine the presence of either centro- 
spheres or centrosomes. If the centrosomes exist as such, 
they would be very difficult to distinguish from the other 
granules of the cytoplasm. 
When the nucleus is about to divide, however, two centro- 
spheres appear, one at either end of the nucleus (Figs. 14, 15). 
They consist of a somewhat dense mass of protoplasm (archo- 
1 I am indebted to the kindness of Professor D. H. Campbell for a supply of this 
material. 
