W. Carruthers — Halonia and Lepidodendron. 149 



surface. But when the specimen has been buried in a sandstone, all 

 trace of the fibre of vegetable substance which once existed has been 

 removed, by the slow combustion that has gone on through this 

 porous rock, and which was impossible in the impermeable plastic 

 material of the shale. In this case we find the cast of the interior as 

 in the first condition described, but lying loose in the mould, and the 

 peculiarities of the exterior exhibited on the mould. Sometimes 

 these exterior markings are the countei-parts of the imbricated leaf 

 bases shown on the fragment, Fig. 3 of the Plate. This is the con- 

 dition of the specimens on which Feistmantel's important observa- 

 tions were based, and which are figured by him. More frequently^ 

 however, the markings in the mould are rhomboid, as in the bifurcating 

 specimen figured by Eichwald (" Letheea Eossica," pi. xi.), and 

 reproduced by Schimper (Traite Paleont. Veg., pi. Ixvi.), and of the 

 principal figure in our Plate. 



To understand the meaning of these two different kinds of external 

 impressions, we must recall the aspect of the surface of ZepidopUoios, 

 which has been shown to be that of Halonia. This generic group 

 was separated by Sternberg from Lepidodendron, because of the long 

 persistent bases of the leaves which clothed the stem after the leaf 

 had fallen ofi'. The difference between the two genera is the same 

 as what is found among the stems of some genera in recent Cycadece. 

 In Cycas the leaves separate at the surface of the stem, vi^hile in 

 Encephalartos the articulation is at some distance from the stem, 

 which is thus permanently clothed v/ith the bases of the leaves. 

 The fall of the leaf is owing to the formation of a transverse layer of 

 cells, which cuts off the petiole by a regular septum, leaving a clear 

 scar of a uniform character in each species. In stems of Encephalartos, 

 which are dead and completely desiccated, the persistent leaf bases 

 may be easily broken away at the place of their attachment to the 

 stem. The scar thus produced is very different from that shown on 

 the outer articulating surface ; it is a rhomb with the margins closely 

 applied to the surrounding scars, and the surface of the scar itself is 

 without any markings except those depending on the accidents of the 

 fracture. 



In Lepidodendron the leaf scar is employed to separate the species, 

 and it appears to supply characters which are constant, and conse- 

 quently of real value, as one would expect from the analogy of recent 

 plants ; but it is obviously necessary to secure the true cicatrix in 

 order to determine these characters. A large number of false species 

 of Lepidodendron have been established, chiefly upon slight differences 

 in the form of the leaf base, without regard to whether the true 

 surface of the scar is being examined or not. Take, for example, the 

 fossils placed in the genus Bergeria, which Presl established to 

 receive Lepidodendroid stems with contiguous rhomboidal leaf scars, 

 perforated at the apex. Unger and others retain this genus, but 

 Schimper places the species at the end of Lepidodendron, as " incertae 

 sedis." None of these fossils with which I am acquainted exhibit 

 any well-marked cicatrix, but are like the scars on the principal 

 stem of Figure 1 of our Plate. The mark of the vascular bundle is 



