PRIMARY STRUCTURE OF CERTAIN PALAEOZOIC STEMS. 361 



C. beinertiana to the genus, though I think highly probable, must be regarded as 

 provisional until the structure of these species is more completely known. 



In the Pitys group the numerous small xylem-strands disposed around a large pith 

 are characteristic. The strands, except the outgoing ones, are more or less deeply 

 embedded in the pith ; as a general rule their structure is mesarch. Here also a single 

 strand passed out through the wood, though it may appear as a double bundle im- 

 mediately below its exit from the pith. The pith is large in all known specimens (1-5 

 cm.)* and may show some approach to discoid structure (P. antiqua), but is probably 

 not really comparable to the Sternbergia pith of the true Cordaitese. The secondary 

 wood has the typical characters of the Pissadendron sub-genus of Dadoxylon, with 

 rather wide elements; the larger medullary rays are always pluriseriate (4-7 cells or 

 more thick). Nothing is known as yet of the cortical structure or of the appendages. 



Lastly, there is the type of Dadoxylon Spenceri. In the one specimen known 

 there are a few small primary xylem-strands scattered at the margin of the pith, and 

 closely applied to the secondary wood. When clearly seen, these strands are found to 

 be of mesarch structure. They pass out in pairs, each pair presumably constituting a 

 single leaf- trace. The pith is of moderate dimensions (5-6 mm. in diameter), and 

 probably not discoid. The secondary wood is once more of the usual Dadoxylon 

 character, but very dense, consisting of small tracheides, with very narrow medullary 

 rays, scarcely ever more than one cell in thickness. In this form — at present isolated 

 — the primary bundles are less conspicuous features than in either of the previous 

 groups. The double leaf-trace is a striking character shared with other stems, both 

 fossil and recent (Ginkgo), in which no such primary strands appear. For these reasons 

 I have not thought it well at present to found a new genus for the D. Spenceri type. 



We have now to consider the probable affinities of these various groups. 



The stems referred to Calamopitys, even on the characters shown in the in- 

 complete specimens by which the British species are represented, are strongly sug- 

 gestive of Cycadofilices, owing to the great development of the primary xylem-strands, 

 and the marked similarity to the structure of Lyginodendron. 



The additional characters present in the Thuringian specimens of the forms described 

 by Count Solms-Laubach seem decisive on this point. The cortical structure, the large 

 leaf-bases, with the characters of Kalymma, the course and structure of the leaf-trace 

 bundles, all point in the same direction. The petioles, known as Kalymma, were no 

 doubt those of compound, fern-like leaves. As mentioned above, it is extremely 

 probable that the petiole, named Rachiopteris multifascicidaris by Mr Kidston, 

 belonged to one or other of our Calamopitys stems. 



The genus Calamopitys has clear affinities with Lyginodendron, but differs in the 

 structure of the petiole, which shows some approach to that of a Myeloxylon. As, how- 

 ever, I have no uew observations to record on this part of the structure, I will pass on 

 to one or two points on which the British species have thrown additional light. It is 



A specimen of Pitys antiqua, with a pith two inches in diameter, is mentioned by Witham, I.e., p. 27. 



