920 MR A. C. SEWARD AND MR A. W. HILL ON THE 



are the same byphal cells in the centre of the stele, the same spirally thickened elements 

 in the medullary rays,* and an identity of structure in the primary and secondary wood. 

 Isodiametric tracheids also occur internal to the primary xylem cylinder, as in the 

 Dalmeny stem.t Unfortunately, the large Arran stems do not afford any positive 

 evidence as to the structure of the leaf-trace after it becomes free from the secondary 

 wood, the tissue in the innermost cortical region having been almost completely destroyed. 

 In one section a leaf-trace with both primary and secondary elements was found close to 

 the edge of the wood, but the fan-shaped group of secondary elements was directed inwards 

 and not outwards as in the Dalmeny stem ; as the trace is isolated and not enclosed in 

 tissue, it has no doubt been twisted, and indeed it may not improbably belong to a 

 Stiornaria and not to the stem with which it is associated. 



The phelloderm of the Arran trunks is identical with that which we have described, 

 there are the same rows of secretory strands as in our stem,}: and in tangential sections 

 of the phelloderm elliptical groups of cells are met with identical with that of PI. I. 

 fig. 8, which we regard as the parichnos.§ Passing to the smaller specimens, we may 

 first mention the specimen represented in pi. xiv. fig. 5 of Williamson's tenth Memoir. || 

 One important fact to notice in connection with this example is the very small amount 

 of secondary xylem in a stem about 12 cm. in diameter. Here, again, there is the 

 closest agreement with the Dalmeny stem as regards the structure of the wood and 

 cortical tissues ; the large sac-like cells shown in our PI. IE. figs. 9 and 15, and PL III. 

 fig. 16, are clearly seen in Williamson's figured section. The middle portion of the 

 cortex in Williamson's specimen presents the characteristic structure which is shown in 

 the parichnos, or portion of the inner cortex accompanying the leaf-trace, represented 

 in our PL I. fig. 8. The small twigs described by Williamson show fairly well-pre- 

 served leaf-cushions, and the form of these, as seen in longitudinal section, suggests the 

 Lepidophloios rather than the true Lepidodendron type. In the smaller axes, 1*2 cm. 

 in diameter, there is no trace of secondary xylem, but each leaf-trace is accompanied by 

 numerous short and spirally thickened tracheids, which in a transverse section of a leaf 

 are seen to be arranged round the strand of elongated scalariform tracheids ; their 

 position and appearance recall the transfusion tissue of many recent plants. IT These 

 short tracheids are well shown also in longitudinal sections of the leaf-trace on the flanks 

 of the tracheal strand. # * The solid xylem strand occupying the axis of the small twig 

 may indicate that these are slender shoots of an Halonial branch ; their stele agrees 



* The thin-walled spirally thickened elements accompanying the leaf-traces in their course through the secondary 

 wood are clearly shown also in Lepidodendron brevifolium, Will. {cf especially No. 499), and in Stigmaria, L. 

 vasculare, etc. 



t Qf. section 448 A. 



X These secretory groups of cells are exceedingly well shown in some sections of the phelloderm of the largest 

 Arran stem in the Binney Collection, Cambridge, as well as in British Museum sections. Similar cell-groups have been 

 observed in sections of L. vasculare ; cf. also Hovelacque (92), p. 58, fig. 17. 



§ E.g., sections 443, 447, 448, 449 in the Williamson Coll. ; cf. also sections of Sigillaria (662 and 665). 



|| Williamson (80). IT E.g., sections Nos. 431, 432, etc. (Williamson Coll.). 



** Cf. section 1955, and several others in the British Museum. 



