208 FLORA OF LOWER COAL MEASURES OF MISSOURL 



especially since .there is doubt as to the generic identity of the snp]30sed 

 fructifications also described under that name. It seems doubtful whether 

 the latter represent any portion of a tree of LepiclopMoios. To the Cannel- 

 ton species probably belongs also the original, which I have not seen, of 

 fig. 4 in the Coal Flora. The points of resemblance and difiPerence in the 

 exposed portion of the compressed bolsters may be noted hj a comparison 

 of Fig. 2, PI. LVI, representing No. 5944 of the Lacoe collection, or 

 PL LVII, which is more compressed, with fig. 4 in the Coal Flora. So, 

 too, fig. 2 of the Coal Flora may with interest be compared with the photo- 

 graph given in PI. LVII of a small portion from a large trunk. This 

 fragment, 42 cm. long, 28 cm. wide, compressed to a thickness of 2 cm., 

 fails to give an adequate idea of the probably gi-eat diameter of the trunk, 

 since neither lateral margin is shown. Before leaving the discussion of 

 the compressed stems we maj note that the gash-like impressions on the 

 decorticated surface of the trunk appear to nearly correspond in position to 

 the ventral trace. The lower vanishing ends of the furrows hardly reach 

 the profile of the upper border of the foliar scars in the compressed speci- 

 mens, while the middle part is slightly below the line of the ventral traces. 

 An interesting feature of the more recent collections from Missouri is 

 the occurrence of isolated or small groups of detached bolsters found in 

 relatively large numbers, especially at Gilkerson's Ford. Several of these 

 are shown in their association with other plants in Fig. Ic, PI. LXI; Fig./, 

 PI. LXII, and Fig. 5, PI. LXIII. The chief details of these may be seen in 

 the photographs. Most of these detached bolsters from this locality appear 

 as casts of the spongy tissue, to which are attached patches of dull carbona- 

 ceous crust, attaining a thickness of about 5 mm., which probably represents 

 cortical tissue as well as the real epidermis. In general it will be noted 

 that the greater inflation of the bolsters is toward or at the side opposite 

 the leaf scar. This side is usually, as shown in Figs. 4 and 7, and PI. LVI, 

 somewhat collapsed and wrinkled, or even slightly infolded. The infold- 

 ing is specially strong near the lateral angles, which in profile often appear 

 quite sharp (Fig. 5). In Figs. 1 and 2, PL LVI, it is plain that the bolsters 

 imbncate for a considerable distance on those below. It also appears that a 

 portion of the dorsal surface of the upper bolster must lie beneath it and 

 upon the upper ventral borders of the bolsters partly underlpng it. If 

 any such expansion exists, however, it must be greatly constricted at the 



