*JounialMaXi?iSo.] Fioycil Microscoj-jical Society. 



149 



witli great care, I took a cast of the wliole of it, and several casts 

 of the separate scars. From them the cbawing of the scar on 

 Plate XLIIL, Fig. 4, has been faithfully produced by Mr. Smith. The 

 original is a natural cast of the outer surface of the stem, whereas 

 Mr. Smith's drawing is from the artificial cast, and consequently 

 exhibits the true aspect of the outer surface of the scar. As is 

 obvious, there are no scales ; but the markings are all produced 

 either by circular pits or by elongated furrows. These are the ends 

 of the vascular bundles, as was determined by Presl and Goppert. 

 The scar is bounded by a distinct boundary line, generally forming 

 a raised margin, as in U. parmatum. This is certainly the boundary 

 of the articulating surface, and shows that the smaller scars scat- 

 tered over the large cicatrix were not the bases of leaves, as supposed 

 by Goppert and Brongniart, but the bundles passing into a branch 

 of some kind, as Presl suggested. 



The difference in the form of the scars on the upper and under 

 half of the stem which has puzzled observers, and has not yet been 

 explained, will be obvious if we consider the structm-e of the stem 

 on which these appendages were borne, as figured and described 

 by me in former numbers of this Journal. On Plate XXXI.* was 

 figured the structure of the axis and vascular cylinder of U. minus. 

 A comparison of this drawing with that of Lepidodendron selagi- 

 noides on Plate XXVII. shows that the tissues of Ulodemlron agree 

 with those of Lcjndodendron as far as they can be compared — 

 and remembering that externally the two genera agree except in 

 the possession of large scars by Ulodendrou — there can be no doubt 

 that the elementary tissues of the one stem can be suppHed from 

 the more perfect specimen of the other. It thus appears that the 

 conical scar has passed far into, if not quite through, the regularly 

 arranged prosenchyma of the cu'cum- 



ference. The vascular bundles, rising pj|j|') o .bia Bi 



upwards and outwards from the cir- 

 cumference of the vascular cylinder, 

 would, in passing into the appendicular 

 organ, penetrate the lower half of the 

 articulating surface at right angles, and 

 would consequently show as circular 

 pits on the cicatrix ; while the bundles 

 on the upper half would penetrate the 

 surface at a very oblique angle, and 

 wouldconsequently show on the cicatrix „. ^.v, o. ,r^ , ^ .,„„. 



1 •' rrn • niagram of the Stem of rlodendron, suovi - 



as more or less elongated furrows. ihlS Ing the direction or the bundles entering 



■ Ti T ,1 -J- c the aerial roots. A, vascular axis; \i. 



Will be apparent by an eXammatlOn OI cylinderofscalariform vessels; Cpareu- 



the accompanying chagram. In species, ''^^'^^ ' ^' pro^^ncuyma. 

 like U. transversum, where the inverted cone of the scar has a de- 

 * ' Monthly Microscopical Joiu-ual,' vol. i. 



