A zolla filiculoides, Lam . 169 
more than two complete coils. Their small size renders 
them unfavourable for a study of the details of development, 
and no attempt was made to study these. 
The ripe prothallium remains completely imbedded in the 
substance of the massula (Fig. 28), and probably the sper- 
matozoids escape by a softening of the outer surface of the 
massula which has a corroded appearance in microtome- 
sections, quite different from the distinct outline of the 
younger ones. 
A comparison of the antheridium with that of other forms 
does not show a very close resemblance to any. From 
Salvinia it differs in the complete surrounding of the sperm- 
cells by the parietal cells, and in the separation of the 
sperm-cells into two groups in the latter. Among the homo- 
sporous Ferns, the antheridium of the Hymenophyllaceae, 
perhaps, resembles it most nearly, especially in regard to 
the arrangement of the parietal cells. In some cases a tri- 
angular opercular cell was observed which, from its position, 
looked as if it had been formed subsequent to the formation 
of the vertical walls, much as in Osmunda. 
Germination of the Macrospores. 
The study of the germinating macrospores involves various 
difficulties. First, to collect a sufficient number it is necessary 
to collect a large number of plants, as each fertile one 
furnishes usually only one or two spores, and only a com- 
paratively small number of plants have them at all. The 
spores only germinate after they have been set free by the 
decay of the indusium, and the best way to get a supply is 
to collect a number of fruiting plants and allow them to 
remain in water until the fertile branches die. These will 
then finally sink to the bottom of the vessel, and may be 
picked to pieces and the spores separated. Spores secured 
in this way will usually germinate promptly, but there is 
considerable difference in this respect ; and, as there is 
nothing to indicate whether or not germination has begun, 
it was only by making repeated sowings and sectioning a very 
