The President's Address. By Duhinfield H. Scott. 139 



for the first time, gave us a clear conception of Palaeozoic Gymno- 

 sperms. While Grand'Eury, with his admirable restorations, re- 

 vealed the habit of these great trees, now only distantly recalled 

 by that of some Southern Conifers allied to the Kauri Pine of New 

 Zealand, Eenault made known the minute structure of every organ 

 of the plant, and afforded a sure basis of comparison with other 

 groups. The result was to show quite clearly that the characteristic 

 Gymnosperms of Palaeozoic times were not, as had been supposed 

 before, members of the Coniferae, but a quite distinct family, com- 

 bining, in various respects, the characters of Coniferae, Cycadaceae, 

 and the family which we now call Ginkgoales, represented in living 

 vegetation by the Maiden-hair Tree alone. While stem and root 

 on the whole present a Coniferous structure, the leaves, apart from 

 their simple form, are almost purely Cycadean ; the very peculiar 

 male catkins (e.g. Cordaianthus Penjoni, above referred to) most 

 nearly suggest those of Ginkgo, with which the female organs have 

 likewise some analogies, though so far as the seed is concerned, a 

 comparison with Cycads is equally admissible (Renault, 1879). 



The investigation of the female fructifications yielded, perhaps, 

 the most striking results of all, for here, in the pollen-chamber of 

 the ovule (in Cordaianthus Grand' Euryi) Renault first found the 

 multicellular pollen-grains to which so much interest attaches, and 

 found evidence that their growth and differentiation went on after 

 they had entered the ovule. To this subject we shall return in the 

 following section. 



The Cordaiteae, this highly developed, but at the same time 

 synthetic Gymnospermous family, extends back in time to the 

 Devonian period, a fact always to be borne in mind by the palaeo- 

 botanist, as showing the enormous antiquity, even of highly 

 organised seed -plants, and thus inculcating caution in our 

 phylogenetic speculations on their origin. Side by side with 

 them, as we now know, lived a great multitude of Fern-like 

 Spermophyta, with apparently far more primitive characters than 

 the Cordaiteae, but not, so far as has yet been proved, of greater 

 antiquity. At the present stage of our knowledge we cannot doubt 

 that the Pteridosperms represent most nearly the earliest type of 

 Seed-plant, but this is a morphological inference, and is not as yet 

 immediately given by the geological record. 



The family of the Poroxyleae, discovered by Renault (Renault, 

 1879), and fully investigated by him in conjunction with Professor 

 Bertrand (1886), is of extreme interest, as tending to connect the 

 Cordaiteae with lower groups. Our knowledge is here, for the most 

 part, anatomical, though M. Grand'Eury has once more stepped 

 in, and is enabling us to recognise these plants in the form of 

 impressions, and, as he believes, to identify their seeds. 



The differences between this group and the Cordaiteae cannot 

 here be fully discussed ; in general vegetative structure there is 



