136 Royal Society : — 



Of this family numerous other modifications are described. Thus 

 TJlodendron and Halonia, very closely allied, if not identical genera, 

 have a structure closely corresponding with that of Lepidodendron 

 Harcourtii, since they possess a very distinct cellular medullary 

 axis enclosed within the ring of medullary vessels, and, besides, 

 exhibit the enclosing ligneous zone at its minimum stage of develop- 

 ment. The remarkable scars of TJlodendron and the tubercles of 

 Halonia appear to have had their most prominent surfaces composed 

 of the true bark-layer deprived of its epidermal bast and parenchy- 

 matous layers, which surround these structures but do not wholly 

 enclose them. These characteristic structures are believed to have 

 supported special organs, into which the epidermal laj'er of the stem 

 has been prolonged, and which the author believes to have been 

 reproductive cones. Favularia corresponds very closely, so far as its 

 cortical layer is concerned, with those already described ; and as 

 Brongniart's Sigillaria elegans is an unquestionable Favularia, the 

 entire series of this subgenus is brought into the closest relationship 

 with the plants described. But the author has further met with 

 some important examples, showing that the stem supported verticils 

 of organs that were neither leaves nor branches, but which are believed 

 to have been cones, thus bringing to light an additional indication 

 of affinity between Favularia, Halonia, and TJlodendron. 



Well-marked examples have also been obtained from the Lan- 

 cashire Lower Coal-measures, the source whence all the specimens 

 described have been obtained, of the outer cortical layers of true 

 Siffillarifs. These specimens demonstrate that the bark of these 

 plants is of the true Lepidodendron type. No example of an un- 

 questionable Sigillaria in which the central woody axis is preserved 

 has yet been seen by the author. 



Stigmaria is shown to have been much misunderstood, so far 

 as the details of its structure are concerned, especially of late years. 

 In his memoir on Sigillaria elegans, published in 1 839, M. Brongniart 

 gave a description of it, which, though limited to a small portion 

 of its structure, was, as far as it went, a remarkably correct one. The 

 plant now well known to be a root of Sigillaria, possessed a cellular 

 pith without any trace of a distinct outer zone of medullary vessels, 

 such as is universal amongst the Lepidodendra. The pith is imme- 

 diately surrounded by a thick and well-developed ligneous cylinder, 

 which contains two distinct sets of primary and secondary medullary 

 rays. The primary ones are of large size, and are arranged in 

 regular quincuncial order ; they are composed of thick masses of 

 mural cellular tissue. A tangential section of each ray exhibits a 

 lenticular outline, the long axis of which corresponds with that of 

 the stem. These rays pass directly outwards from pith to bark, and 

 separate the larger woody wedges which constitute so distinct a 

 feature in all transverse sections of this zone, and each of which 

 consists of aggregated laminae of barred vessels disposed in very 

 regular radiating series. The smaller rays consist of vertical piles 

 of cells, arranged in single rows, and often consisting of but one, 

 two, or three cells in each vertical series ; these latter are very 



