9 6 
Agnes Robertson. 
to hear in mind that the evolution of the microsporophylls of the 
Gymnosperms is still very obscure; there are great gaps between 
the microsporophyll of the Cycad, with its numerous clustered sori 
on the lower surface,—the pinnate synangia-bearing microsporophyll 
of Bennettites (Wieland, 25, Fig. 87 p. 164),—the erect stalked 
cluster of pollen sacs of Cordaites, —and the stamen of any of the 
Coniferales. 
The minute structure of the reproductive organs does not give 
much positive evidence for the idea that the group is a primitive 
one, except in a single point to which I am disposed to attach some 
importance. This is the mode of development of the proembryo 
(see Coulter and Chamberlain, 13). In the Cycads and Ginkgo and 
probably Bennettites (Wieland 25, p. 125 and PI. XXX.) the fertilised 
egg nucleus divides repeatedly before any walls are formed ; but in 
Pinus as soon as four nuclei are produced they become separated 
by walls. Taxus and Cephalotaxus recall the Cycadean condition 
in that no walls are formed until at least sixteen free nuclei have 
arisen by division from the fusion nucleus. 
The mature embryo of the Taxacere has two cotyledons ; a 
similar dicotyledonous condition prevails throughout the Cycads, 
Bennettites , and Ginkgo, while polycotyledony is characteristic of the 
Abietineae. 
A further point in favour of the view that the Taxoideae are a 
relatively primitive group is the resemblance which exists between 
the female “ flowers” of Taxus and Cordaites. To this point I will 
return later. 
The most important arguments brought forward by those who 
believe that the Taxoideae are specialised rather than primitive are 
concerned with the structure of the megaspore and the female 
gametophyte. Dr. R. B. Thomson of Toronto has recently drawn 
attention to the “ megaspore membrane ” of the Gymnosperms (21). 
In common with other authors he regards the Gymnosperms as 
ultimately descended from forms in which the megaspores were set 
free from the sporangium ; in such cases a firm protective spore 
coat would obviously he of use, but it would lose its function in the 
later stages of evolution when the megaspore became embedded in 
the nucellus. Hence Dr. Thomson looks upon the megaspore 
membrane of the Conifers as a vestigial structure, retained in the 
more primitive forms and gradually lost in the more advanced. He 
finds that the megaspore membrane is present, associated with a 
tapetum which he regards as of sporogenous origin, in the Cycads, 
