LYCOPOUIALES— LEPIDODENDRE.E— LEPIDOPHLOIOS. 207 



In PI. LVII, and PI. LVI, Figs. 1 and 2, I have shown several specimens 

 which iUustrate the general characters of the compressed stems of the species 

 before me. As seen in the enlargements, PI. LVI, Fig-. 2a, 2h, and PI. LVII, 

 Fig. la, the exposed surface is rhomboidal, the sides being fairly straight, 

 the lateral wings very acute and often prolonged and sinuate. The lower 

 angle is rounded at the median line, while the borders on either side are 

 usually slightly concave, though sometimes very nearly straight. The leaf 

 scar, as described above, is remarkable for the prolongation of the lateral angle, 

 its great width as compared with its altitude, the rather broadly rounded base, 

 and the generally flat or often distinctly emarginate upper angle. 



In the form of the compressed cushions and of the leaf scars, our speci- 

 mens agree so closely with fig. 2, pL cv, of the Coal Flora as to suggest that 

 both fragments might have been found in the same locality. Furthermore, 

 this suggestion is emphasized by the fact tliat in No. 6943 of the Lacoe col- 

 lection, which was marked by Professor Lesquereux as the original of fig. 2 

 in the Coal Flora, and which came unquestionably from Cannelton, as was 

 originally stated, the lower borders of the bolsters are generally verj^ much 

 more rounded than is shown in the figure. A comparison, however, of the 

 cicatriciiles and ventral trace shows that in the specimens from Missouri 

 the central cicatricule or vascular scar is generally distinctly below the 

 middle of the leaf scar, the lateral cicatricules being very close to the basal 

 margin, and the ventral trace close to the upper border of the scar, while in 

 the figure in the Coal Flora the vascular trace is close to the upper margin 

 of the scar, the lateral traces being situated about midway in the altitude 

 of the scar, while the ventral trace is generally nearer the borders of the 

 bolster next above. In bolsters of about the same size the ventral trace is 

 seen to be only about one-half as far above the leaf scar as in the type 

 from Cannelton. Moreover, while the cicatricules on the type of fig. 2 are 

 often obscure, there seems to be good reason for representing them as is 

 done in the figure. Finally, the real form of the uncovered bolster in the 

 original from Cannelton shows a very much roiinder type, its form being 

 nearly intermediate between L. Van Ingeni and L. aiiriculatus Lx. Hence, 

 notwithstanding the close resemblance in the form of the exposed part of 

 the bolster and leaf scar in the type of fig. 2 and in the Missouri tree, I 

 feel constrained to regard them as belonging to different species. For the 

 Cannelton type of bark the name LepidopJiloios dilatatus may be retained, 



