1871.] Prof. Williamson on Lepidodendra and Sigillarise. 503 



these latter are very numerous and intervene between all the numerous 

 radiating lamina? of vessels that constitute the larger wedges of woody 

 tissue. The vessels going to the rootlets are not given off from the pith, 

 as Goeppert supposed, but from the sides of the woody wedges bounding 

 the upper part of the several large lenticular medullary rays, those of 

 the lower portion of the ray taking no part in the constitution of the 

 vascular bundles. The vessels of the region in question descend vertically 

 and parallel to each other until they come in contact with the medullary 

 ray, when they are suddenly deflected, in large numbers, in an outward 

 direction, and nearly at right angles to their previous course, to reach the 

 rootlets. But only a small number reach their destination, the great 

 majority of the deflected vessels terminating in the woody zone. A very 

 thick bark surrounds the woody zone. Immediately in contact with the 

 latter it consists of a thin layer of delicate vertically elongated cellular 

 tissue, in which the mural tissues of the outer extremities of the medullary 

 rays become merged. Externally to this structure is a thick parenchyma, 

 which quickly assumes a more or less prosenchymatous form and becomes 

 arranged in thin radiating laminae as it extends outwards. The epidermal 

 layer consists of cellular parenchyma with vertically elongated cells at its 

 inner surface, which feebly represents the bast-layer of the other forms 

 of Lepidodendroid plants. The rootlets consist of an outer layer of 

 parenchyma, derived from the epidermal parenchyma. "Within this is a 

 cylindrical space, the tissue of which has always disappeared. In the 

 centre is a bundle of vessels surrounded by a cylinder of very delicate 

 cellular tissue, prolonged either from one of the medullary rays or from the 

 delicate innermost layer of the bark, because it always accompanies the 

 vessels in their progress outwards through the middle and outer barks. 



The facts of which the preceding is a summary lead to the conclusion 

 that all the forms of plants described are but modifications of the Lepidoden- 

 droid type. The leaf-scars of the specimens so common in the coal-shales 

 represent tangential sections of the petioles of leaves when such sections are 

 made close to the epidermal layer. The thin film of coal of which these 

 leaf-scars consist, in specimens found both in sandstone and in shale, does 

 not represent the entire bark as generally thought, and as is implied in the 

 term " decorticated" usually applied to them, but is derived from the 

 epidermal layer. In such specimens all the more central axial structures 

 (viz. the medulla, the wood, and the thick layer of true bark) have disap- 

 peared through decay, having been either destroyed or in some instances 

 detached and floated out ; the bast-laver of the epiderm has arrested the 

 destruction of the entire cylinder, and formed the mould into which 

 inorganic materials have been introduced. On the other hand, the woody 

 cylinder is the part most frequently preserved in Stigmaria ; doubtless 

 because, being subterranean, it was protected against the atmospheric 

 action which destroyed so much of the stem. 



It is evident that all these Lepidodendroid and Sigillarian plants must 



