299 
investigations upon the seeds of the Cycadales, Ginkgoales, 
Coniferales and Gnetales in connection with the insight 
we get here about the integumental origin. 
The seeds I have described, and which represent the 
most characteristic specimens of the discovered fossil seeds, 
are easily to range in a series, suggesting a possible line 
of evolution. [| do not mean to say that they have origi- 
nated from an other, but that the tendency of investing 
the macrosporange, has developed in one seed a higher 
and more differentiated and specialized organ than in the 
other. So that the supposition arises that the highest 
developed organ has not been formed all at once, but 
has passed through the lower stages. 
The most primitive stage is probably represented by 
Physostoma. The testa clearly indicates an origin from 
several segments, which are entirely free at the top, and 
though connate at their base, they also show there dis- 
tinctly their original separation. À differentiation of the 
tissue has not yet taken place, but an attempt for more 
protection has been disclosed by the formation of a thick 
hair covering at the outerside. 
Lagenostoma and Sphaerostoma are higher developed. 
In both the units, composing the integument are more 
closely fused and are recognizable only by the ring of 
loculi around the micropyle, and the vascular bundles, 
ending in them. À cupule is present in both seeds and 
probably in connection herewith the differentiation of the 
testal-tissue is not much advanced. In Lagenostoma a 
palisade layer is present, corresponding with the hypo- 
derma in Sphaerostoma, which consists of a layer of scle- 
rotic cells. These layers are apparently the first indications 
of the later sclerotesta. 
Finely Trigonocarpus and Mitrospermum have reached 
the highest degree amongst the Pteridospermae. In the 
integument of both there is a much slighter markation of 
Recueil des trav. bot. néerl. Vol. XVII. 1920. 20 
