OF THE FOSSIL PLANTS OF THE COAL-MEASTTEES.

515

extended than is the case in the conjugating cells of the Phycomyces nitens describedby the French biologists.

That these dark-coloured, hollow, thin-walled branching tubes are as different aspossible from the transparent and colourless siliceous spines of the Radiolarians is tooobvious to require further remark. Soft and flexible in their young state, they becamebrittle only when more matured, a condition to which I have referred in an earlierpart of this memoir as not uncommon amongst the macrospores of such recentLycopods, as Selaginella Martensii. The cells seen in figs. 42 and 45 are wholly^indistinguishable from similar endospermic cells seen in the Lycopodiaceous macro-spores figured in Plate 23 of my last memoir, Part IX. Hence I adhere to mypreviously expressed conviction that the Traquairce are really vegetable organisms,and that there are strong grounds for supposing them to be Cryptogamic macrospores.Professor Strasburger suggests that their nearest allies will possibly be found inthose of Azolla and other Rhizocarpous genera.

In my last memoir I gave small figures ( loc . cit., Plate 23, figs. 72, 73, and 74) ofthree small bodies, respecting which I observed : “ It is impossible to overlook thestriking resemblance of these little objects to the fossil Xanthidia of the chalk flints,and to the zygospores of some of the Desmideae.” When these words were pennedI was not certain that these objects might not prove to be young states of some of thenumerous spores with which the Halifax rock abounds. Since then I have obtainednumerous additional examples of these objects in sections for which I am indebtedto Messrs. Spencer and Earnshaw, and find their characteristic features to be soconstant that I cannot doubt their being matured organisms, whatever may be theirbotanical nature. That they were all more or less spherical, with radial appendagesdistributed over their entire periphery, is certain.

Fig. 51 represents one of these ^objects in their most common aspect. The diameterof the central disk is about ’0014. The radiating arms are of somewhat variablelength. These arms always branch more or less peripherally as represented by thefurther enlarged fig. 52. It is not always easy to trace their exact ramifications owingto imperfections in their mineralisation, but in those which are well preserved thereare usually two or three primary divisions, the extremities of which are further sub-divided.

Fig. 53 represents another of these objects, the extreme diameter of the disk of which,exclusive of the radiating arms, is ’0018. The arms in this example are rather shorterthan in the last one. The section has here passed tangentially through the upper-most portion of the disk—hence we see at a the bases of arms springing at regularintervals from what remains of its convex surface. Fig. 56 is a less highly magnifiedfigure of a much larger specimen, the disk of which has a diameter of '006. It ischiefly interesting from the fact that it unmistakably exhibits an inner structurelessmembrane, b, devoid of all radial extensions. I have found faint evidences of theexistence of such a membrane in several of my specimens, leaving no room for doubting