_p.c_, Pollen chamber in the nucellus _n_, which is fused to the coat _c_ to the level _l_; _sc_, stony layer in coat; S, the big spore, filled with endosperm tissue (in this case green in colour); _e_, egg cells, one of which will produce the embryo after fertilization.] The nature of the male cell in _Cordaites_ is not yet known, but there is reason to suspect it may have been actively swimming also. As this is uncertain, however, we may consider _Ginkgo_ the most highly organized plant which has such a primitive feature, a feature which is a bond of union between it and the ferns, and which, when it was discovered about a dozen years ago, caused a considerable sensation in the botanical world. To turn now to the fossil records of this family. Leaf impressions of _Ginkgo_ are found in rocks of nearly all ages back even to the Upper Palæozoic. They show a considerable variety of form, and it is certain that they do not all belong to the same _species_ as the living plant, but probably they are closely allied. Fig. 69 shows a typical impression from the Lower Mesozoic rocks. In this specimen, the cells of the epidermis were fortunately sufficiently well preserved to be seen with the microscope, and there is a distinct difference in the size and shape of the cells of living and fossil species, see fig. 70; but this difference is slight as compared with the great similarity of form and appearance, as can be seen on comparing figs. 69 and 66, B, so that the fossil is at the most a different species of the genus _Ginkgo_. Among the fossil leaves there is greater variety than among the living ones, and some which are very deeply lobed so as to form a divided palm-like leaf go by different names, e.g. _Baiera_, but they are supposed to belong to the same family. Fossil seeds and male cones are also known as impressions, and are found far back in the Mesozoic rocks. From the fossil impressions it is certain that _Ginkgo_ and plants closely allied to it were very widespread in the past, as they are found all over Europe as well as the other continents. Particularly in the Lower Mesozoic rocks _Ginkgo_ seems to have been a world-wide type growing in great abundance. [Illustration: Fig. 69.—Leaf Impression of _Ginkgo_ from Mesozoic Rocks of Scotland] [Illustration: Fig. 70.—Showing Epidermis with Stomates from the lower side of the Leaf seen in fig. 69 _e_, Epidermis cells; _s_, stomates; _v_, long cells of epidermis lying over the veins.] In the Palæozoic the records are not so undoubted, but there is strong evidence which leads us to suppose that if the genus now living were not then extant, at least other closely related genera were, and there seems to be good grounds for supposing that _Ginkgo_ and _Cordaites_ may have both arisen from some ancient common stock. CHAPTER X PAST HISTORIES OF PLANT FAMILIES III. The Bennettitales