proximated warts, which are somewhat more pressed together. This difference appears to me of no importance, as Knorria imbricata (of which Knorria acicularis is perhaps only a variety) also occurs with more or less approximated warts. Schimper, who set much value upon this difference, has separated the species as Knorria Bailyana. Opinion is known to be much divided as to the relation of Knorria to Lepidodendron. Goppert, in his first work on the plants of the transition rocks, made twelve species of Knorria, but afterwards reduced them to one, placing most of them with Lepidodendron Veltheimianum ; whilst Geinitz and Schimper hold Knorria imbricata to be a distinct plant. The numerous specimens from Bear Island have inclined me to this view, which I have dwelt upon more fully in my ' Flora of Bear Island.' It is true that Schimper connects with Knorria Bailyana (with a query) the fruits which I refer to Cyclostigma, and also (again with a query) the Cyclostigma minutum, Haught., of which he has only received very fragmentary pieces. It may, indeed, be a question whether our Knorria is not a peeled Cyclostigma. This does not appear to me very probable — because, in the specimens figured by me in figs. 2 and 3, single portions are peeled ; under this bark, however, we by no means have Knorria. We see, indeed, in fig. 3, the impression of the inner bark, but no Knorria-structure, and just as little in fig. 2. That it is not only the bark which we have here, but pieces of the whole stems and branches, we see from the thick masses of coal, and because in some places the impression of the warts at the back is to be seen.
4. Lepidodendron Veltheimianum, Sternberg.
The specimen represented in fig. 1 is from Tallowbridge ; but I received a similar one from Kiltorkan. It is only the impression of a short fragment of a branch ; and as in this the leaf-scars are not so beautifully and sharply defined as in the stems, the exact determination is very difficult. We recognize upon the branch very closely approximated leaf-scars, which touch each other at the margins. They are rhomboidal, longer than they are broad. Inside them we have an oblong or obovate depression, pointed at the base (magnified in fig. 1 b) . The cicatricula is not preserved. In their form and thick-set position these leaf-scars resemble those of the branches of L. Veltheimianum ; but a more certain determination cannot be made from the present specimens.
I received this specimen as Lepidodendron elegans, Haught. A L. Griffithii, Brongn., is not known to me. As far as I know, Brongniart has nowhere described a species under this name. The formation of the bark is so very different from that of Cyclostigma, that it cannot be a branch of the same tree. On the contrary it may be a question whether the Cyclostigmata may not be roots or, more properly speaking, rhizomes of Lepidodendron. It may be alleged that they remind us forcibly of the Stigmarioe, which are now pretty generally considered to be rhizomes or root-formations, and connected either with Sigillaria, Lepidodendron, or Knorria. So long, however, as no traces are found of such a connexion between Cyclostigma
