56 Proceedings of the Royal Physical Society. 



ments twisted in a spiral manner, and he was successful in 

 discovering specimens which showed that these curious plants 

 grew in verticils or umbels. 



Pala^ohromelia was supposed to differ from Palmoxyris by 

 the body of the fossils of the former genus being destitute of 

 the scales supposed to occur on the species of Palceoxyris. 

 The supposed difference, however, between Palmohromelict 

 and Palceoxyris is not one of structure, but of preservation, — 

 for in the former had been placed nothing other than un- 

 compressed examples of Palceoxyris. 



Schimper has pointed out the identity of these two genera,^ 

 and under the name of Spircmgiuni, describes them as com- 

 posed of several leaves (or carpels ?), which are produced at 

 their base, and united to form a hexagonal petiole, which 

 decreases insensibly towards its lower extremity ; above this 

 petiole the leaves (or carpels ?) are twisted in an ascending 

 spiral of 1 or IJ turns around an axis of apparently little 

 consistency or thickness, on which had, perhaps, been fixed 

 the floral organs or seeds. Above the fusiform swelling, 

 produced by the tortion of these leaves united by their 

 margins, the valves become straightened and prolonged in 

 linear appendages, which are united to form a straight, more 

 or less elongated rostrum, resembling a petiole, free or 

 flexuous. This multivalved capsule, elliptic or fusiform, and 

 twisted as those of Loasa, Relictcres, and many Orchids, pro- 

 bably opened by dehiscence. The valves had little thickness, 

 and their lines of union formed the carinated ridges, whose 

 crossing each other, produced in consequence of pressure, the 

 regular rhomboidal impressions which some have mistaken 

 for scales. The discovery of impressions less compressed, 

 and of complete moulds, has shown that these scales do not 

 exist, but that they are caused by the twisting of capsules 

 or involucres in a spiral manner, of which the anterior, 

 crossing the posterior circumvolutions, form, when com- 

 pressed, the more or less sharp angles (of the supposed rhom- 

 boidal scales), according as the spires are more or less upright 

 in their ascent. 



^ Traite d. paleont. veget., vol. ii,, p. 514. 1872. 



