i 78 Beer . — On the Development of the Spores of 
increases rapidly in bulk, whilst the cells surrounding the sporogenous 
mass, to the extent of several layers, assume the character of a tapetum 
which gradually becomes disorganised ; finally the sporogenous cells 
separate and the majority of them divide into tetrads ; but a considerable 
portion of them, scattered throughout the sporogenous mass, become 
disorganised without undergoing division ; this is similar to what has 
been observed in the case of Ophioglossnm V 
My own observations begin with a stage at which the sporogenous 
cells have reached their full number, but in which the tapetum has not 
yet become disorganized (PI. XII, Fig. 15). The tapetal layer is composed 
of radially elongated cells many of which have divided once or, more rarely, 
twice by periclinal divisions (PL XI, Fig. 1, and PI. XII, Fig. 15). In 
Botrychium Cardiff observed periclinal divisions to follow one another until 
the tapetum became four or five cells in thickness. Stevens, however, does 
not confirm this statement and finds the tapetal layer of B. virginianum 
to be, as a rule, two cells in thickness, which would be more nearly in 
accordance with what I have seen in H elminthostachys. 
The membranes of the tapetal cells are delicate, but clearly give the 
reactions of both cellulose and pectose. In this feature the tapetum of 
H elminthostachys apparently differs from that of Botrychium , in which 
Stevens found each tapetal cell to be ‘ deliminated by a plasmatic 
membrane merely/ Although Cardiff does not actually describe the 
cell-walls of the tapetum, his words certainly imply the occurrence of a 
cell-wall other than a plasmatic membrane. 
The sporogenous cells which lie within the tapetal layer are mostly 
square or oblong cells which measure about 12-18 /x across (Figs. 1 6 
and 17). Their walls give the characteristic pectic reactions, but neither 
cellulose nor callose is to be found in them. The cytoplasm of these 
cells is fairly dense and contains a number of plastids usually crowded 
with starch. In slightly older sporangia the tapetal walls begin to break 
down, and in these disorganizing membranes I no longer succeeded in 
obtaining the cellulose reactions, although pectose was still clearly demon- 
strable 1 2 . 
The cytoplasm derived from the disorganized rells flows together 
(Figs. 2 and 16) and increases in amount, whilst the nuclei, which are 
mostly gathered together in little nests or groups at the inner or outer 
periphery of the tapetal layer, have become more numerous. The groups 
of nuclei are no doubt partly formed by the approximation of nuclei from 
several disorganized cells, but there can be little doubt that they also 
represent, to some extent, a number of daughter-nuclei which have origin- 
1 Spore-producing members, Part II, 1896, p. 35. 
2 I state this fact for what it is worth, although the failure of the cellulose reactions may be due 
to the increased difficulty of demonstration rather than to the real absence of this substance. 
