STRUCTURE AND AFFINITIES OF A LEPIDODENDROID STEM. 921 



closely with the small vascular strands given off from the stele of the larger Halonia 

 described by Williamson in 1883. In the small twigs the two arms of the parichnos, 

 composed of large thin-walled cells, are found in close proximity to the leaf-trace 

 bundles, as seen in the fleshy leaf-bases.* In the older stems from which the leaves 

 have fallen, and in which a considerable thickness of phelloderm has been developed, 

 the leaf-trace becomes disorganised on its way through the outer cortex, and the 

 parichnos strands lose to some extent their regularity of arrangement ; their cells, too, 

 appear to be somewhat compressed by the pressure of the mass of phelloderm, through 

 which the parichnos is passing. The cells in the parichnos figured in PI. I. fig. 8 are 

 rather smaller and more compact than those in the parichnos arm of a leaf-base. 



The Williamson Collection includes one specimen of a piece of rock from Arran 

 (No. 62) which shows imperfect surface markings in the form of lozenge-shaped depres- 

 sions, which no doubt represent the leaf-bases of a Lepidodendroid plant, but the pre- 

 servation is far from satisfactory, and we have no proof of the connection between this 

 cast and the petrified stems. 



There is a specimen of Lepidophloios Wilnschianus in the fossil-plant gallery of the 

 British Museum, from the Calciferous sandstone of Craigleith (No. 52,625), which repre- 

 sents a stem with two steles 1 cm. in diameter ; the stem has been cut across trans- 

 versely just below the point at which the two branches of the bifurcating axis became 

 free. It is interesting as showing the absence of secondary wood, although the section 

 measures 19 by 14 cm. ; this is one of the characters shared by L. Wilnschianus and 

 the stems known as L. Harcourtii. 



We may next consider the question as to the possible identity of the Arran and 

 Dalmeny stems with plants described under other specific names. There are many 

 points of resemblance between these stems and the diploxyloid axes from Burntisland, 

 named by Williamson Lepidodendron brevifolium,f but a comparison of the structural 

 features does not lead us to regard the latter form as specifically identical with the 

 Arran and Dalmeny species. There are various characteristics shared by Williamson's 

 species, Lepidophloios fuliginosus (Will.), and L. Wilnschianus (Will.), but the former 

 species is no doubt distinct ; the nature of its secondary xylem and other features prove 

 it to be a well-defined specific type. 



The most interesting comparison | is with the species long known as Lepidodendron 

 Harcourtii (Witham). The plant bearing this name was first described and figured by 

 Witham,§ and afterwards by Lindley and Hutton in their Fossil Flora || ; subsequently 

 more accurate figures were given by BrongniartH in 1837 and 1839. 



* Of. sections 428, 443, etc. 



t Williamson (72), p. 310. In the index (Pt. II.), published in 1893, Williamson speaks of this species as 

 Lepidophloios (p. 10). 



X For the opinions expressed as to the possible identity of the Dalmeny plant and L. Harcourtii, one of us must be 

 held responsible (A. C. S.). 



§ Witham (32) and (33), p. 51, Pis. xii. and xiii. || Lindley and Hutton (33), vol. ii. pis. xcviii. and xcix. 



IT Brongniart (37), pis. xx. and xxi. ; (39), pis. xxx. and xxxi. 



