On the Structure of Lepidodendron selaginoides. 99 



It will be necessary, before re-describing this plant, to give 

 an outline of its structure as generally stated, and as the 

 specimen to which these remarks specially refer seems similar 

 to that described by Professor Williamson, it will suit our 

 purpose best to quote his own words : " The medullary axis 

 consists of a very peculiar admixture of barred cells, and 

 vessels also barred. I abstain, as I have done in my previous 

 memoir on calamites, from designating these vessels as scalari- 

 form, because I have not yet found them to be thickened at 

 their angles with continuous deposits of lignine, as is the 

 case with the true scalariform vessels of ferns. . . . The 

 cells of this structure, in the specimen figured, exhibit a ten- 

 dency to diverge into two forms. We have one thick walled 

 series, arranged in vertical rows, the transverse septse of 

 which are sometimes rectangular in relation to their sides, 

 but much more frequently oblique, the obliquity tending 

 sometimes in one direction and sometimes in another, even 

 in the same pile. The sides and ends of these cells are alike 

 richly barred. Sometimes the bars are regularly parallel 

 with each other, and arranged transversely, as in the vessels ; 

 but very frequently they describe a series of curves, as if one, 

 two, or even three of the angles of the cells had been centres, 

 from which corresponding series of concentric segments of 

 circles had been drawn. . . . The cells of the other class 

 are much smaller, have very thin walls, and appear to be 

 small masses of ordinary parenchyma intermingled with the 

 other medullary tissues. It is possible, but not probable, 

 that this difference is due to mineralisation. . . . The 

 vessels are often almost undistinguishable from sections of 

 the barred cells ; indeed, we appear to have here strong evi- 

 dence of their primarily cellular character. In the specimen 

 figured, those of the centre of the medulla are somewhat 

 widely separated by the two kinds of cellular tissue ; but 

 this separation only extends over a small area. In the 

 peripheral portions of the medullary axis they are closely 

 conjoined, the cellular element becoming less abundant, 

 especially the delicate parenchymatous tissue, which is so 

 much more copious in the centre of the structure. 



" Immediately investing the medullary axis is a thin cylin- 



