630 Kershaw. — Structure and Development of the 
walls between them. The megaspore membrane forms a conspicuous 
covering, but is thin in comparison with that of other Cycads, and shows no 
differentiation into layers as described for some genera, e. g. Cycas revoluta } 
In the oldest ovule examined the megaspore membrane is thin and un- 
differentiated. If, as Thomson suggests, a thick megaspore membrane is to 
be regarded as a sign of primitiveness, then in that respect Bowenia as com- 
pared with other Cycads must be regarded as well advanced. The sporo- 
genous tissue which formed so large a mass round the embryo sac of the 
younger ovule has almost entirely gone. It has evidently failed to keep 
pace with the advancing embryo-sac, and appears as one or two layers 
of granular, compact cells at the outside and a few broken and disorganized 
cells lying loosely around the megaspore membrane (Fig. 8, PI. LXI). 
There is no sign of any active formation of new cells. Evidently at a 
slightly earlier stage the formation of new cells ceased, then, as is shown in 
Fig. 8, PI. LXI, the cells of the tissue separated from one another, gradually 
dissolved, and were absorbed by the encroaching embryo-sac ; cf. Zamia . 1 2 
In rather later stages, when the prothallus tissue fills the embryo-sac, 
the whole of the nutritive tissue has disappeared. The length of time which 
this tissue persists and its appearance during the last stages varies in different 
genera. In an ovule of Stangeria , with the pollen-chamber beginning 
to form and the embryo-sac completely filled by the prothallus tissue, 
the sporogenous tissue is shown 3 to still persist in the form of a single layer 
of much enlarged cells with smaller crushed cells on their inner side. These 
are the remains of a tissue which earlier appeared as a compact surrounding 
mass very similar to that in the young stages of Bowenia . The embryo-sac 
gradually absorbed the sporogenous cells around it, except the outermost 
layer, which enlarged to form the definite layer which Lang termed a 
tapetum. How long this layer persists cannot at present be stated. Lang 
was unable to follow it, for in the fertilized seeds which he had all traces 
had disappeared. A further supply of Stangeria material which I have 
examined confirms Lang’s description, but does not help in this point. 
In an ovule of Dioon 4 with prothallus and pollen-chamber development 
a definite ‘ jacket layer ’ is present round the prothallus, and this acts 
as a tapetum. It thus appears from the present facts that the sporo- 
genous tissue of Stangeria and Dioon persists much longer than in either 
Boivcnia or Zamia? 
The integument of the ovule at this stage shows indications of the 
three layers characteristic of Cycads. The cells destined to form the 
stone layer are smaller than those forming the fleshy parts, though they 
1 Thomson, R. B. : The Megaspore Membrane of the Gymnosperms. Univ. Toronto Biol. 
Series, No. 4, 1905. 2 Smith, F. G., loc. cit. 3 Lang, loc. cit. 
4 Chamberlain : The Ovule and Female Gametophyte of Dioon. Bot. Gaz., vol. xlii, 1906. 
5 Smith, loc. cit. 
