301 
Sphaerostoma a plane of dehiscence is formed between 
cupule and integument, preparing the shedding of the 
seed, and in Gnetopsis (135) a similar envelopment en- 
closes two to four seeds!), this may be a proof for the 
supposition that the cupule is not homologous with an 
outer integument. Àn integument is always a permanent 
investment around the nucellus, therefore the cupule may 
be regarded somewhat as a modified sporophyll or a part 
of it. The protecting influence, of the cupule, during the 
connection of the seed to the motherplant is well pro- 
nounced by the lack of differentiation of the integument. 
The free position of the nucellus within the testa is a 
primitive state, mostly the case in the young ovules also. 
By a subsequent growth in the chalazal region, the fusion 
of nucellus and integument is accomplished. This coherence 
in the fossil seeds is difficult to control, as the soft tissues 
of the nucellus have loosened and are collapsed. In all 
cases the top remains free and then forms the lageno- 
stome or pollenchamber. 
À real micropylar tube is not yet formed in Physostoma, 
because the tentacles have not closed and fused. In 
Lagenostoma and Sphaerostoma also it is still less deve- 
loped, the nucellar-apex being on the same level with the 
surrounding lobes of the canopy. Àt first in Trigono- 
carpus and Mitrospermum the tube becomes longer and 
also in this respect we get more the gymnospermous type. 
À second group of plants, to which we now have to 
pay attention are the Cycads. Amongst them the genus 
Cycas is distinctly characterized as concerns the place 
and the structure of the ovule. À single sporophyll is still 
present here, whereas in the other genera a strobilus 
1) The seed of Gnetopsis 1 have not described, because the mentioned 
feature excepted, it is not much different from Lagenosfoma and 
Sphaerostoma. 
