GYMNOSPERMtE. 



115 



Williamson as Astrowyelon. It appears that the stem and branches grew together in 

 exactly the same relation as those of an ordinary exogenous tree, the branching not 

 differing materially in its outward appearance from that of a Pine. He considers it a 

 Cryptogam whose affinities were possibly with Marsilea ; and that the large radiating 

 lacunae of the bark show that it was at least semi-aquatic in its habits. 



Contemporaneous with these were many varieties of truly gymnospermous stems. M. 

 Renault finds that the wood of Toroxylon is dotted with areolated puncta similar to those 

 distinguishing the spiral vessels of Cycads and Araucarias; and Sir W. Dawson has described 

 no less than five species of the Coniferous Badoxylon from the American Middle Devonian. 

 Prof. Williamson long ago demonstrated that the supposed British plants known as Stern- 

 hergice were merely inorganic casts of the hollow discoid pith of a Badoxylon, and he has 

 more recently shown that in some of these Badoxylons, double foliar bundles pass off to 

 each leaf-petiole, as in the recent Ginl~yo or Salisburia. That the Gymnospermse had already 

 attained a considerable numerical development in the Carboniferous Period is still more 

 apparent from the seeds met with in certain localities. The affinities of Triyonocarpus with 

 the drupaceous seeds of the existing Ginkgo were pointed out by Sir J. D. Hooker and 

 Mr. Binney as long ago as the year 1855.^ Mr. Carruthers, in 1872, figured two 

 species of Cardiocaqjon attached to their axes,^ and clearly pointed out their 

 Gymnospermous character. The fortunate discovery by M. Grand'Eury of a number 

 of silicified seeds at St.-Etienne and Autun; and the magnificent posthumous work 

 upon them of M. Ad. Brongniart, completed by M. Renault, has shed a flood 

 of light upon the subject. These seeds have been placed in a number of genera, 

 and are of many and diverse forms, but as yet it has not been possible to allocate them 

 definitely among the previously known Carboniferous genera. Prof. Williamson, in 

 describing a large series of similar seeds from the Carboniferous deposits of Lancashire 

 and Burntisland, seems to have experienced the same difficulty.^ Some of them are 

 more complex than those of many existing Gymnospermous seeds, those provided with 

 a testa and a double membrane recalling seeds of Cycads and Taxese, especially Ginkgo. 

 Notwithstanding their divergence of form, from simply bicarinated to a structure com- 

 posed of many radiating elements disposed round a common axis, they have one peculiarity 

 in common, the possession of a cavity at their micropylar extremity, called the " chambre 

 pollinique " by Brongniart, and the " lagenostome ^' by Prof. Williamson. Pollen grains, 

 sometimes much larger than those of existing Gymnosperms, entered this chamber 

 through the micropyle, increasing in size during their stay in it and developing septa or cell- 

 walls internally. Underneath this, the lagenostome or chamber, is the albumen or endo- 

 sperm. Germination has not taken place in any of the fossil seeds and the development 

 of the embryo is unknown. The great importance of these seeds lies in the fact that 



1 ' Phil. Trans.; vol. cxlv, part 1, p. 149. 2 ' Geol. Mag.,' vol. ix, pp. 55—57, figs. 1—3. 



3 " Eighth Memoir on the Organisation of Fossil Plants of the Coal-Measures," ' Phil. Trans.,' vol. clxvii, 

 1877, p. 213. M. Lesqnereux has figured a number of similar seeds from the American Carboniferous. 



IG 



