LINNEAN SOCIETY OF LONIJOX. 21 



Sigillaria under Ferus, -w'hile he included Lindley's genus Caido- 

 pteris (true Tree-ferns) under Sigillaria, recognising that the 

 fern-affinities of C'aulo^yteris were more evident than those of 

 the Sigillarice proper. Brongniart shows that the Sigillarias 

 could not have been Cacti, as Martius, or Euphorbife, as Artis 

 had supposed, for they were not succulent, but woody plants. 

 He ai'gues that they could not have been Dicotyledons at all, 

 for their form gives no indication of growth in thickness at tlie 

 base of the stem. In view of subsequent developments, his 

 rejection of growth in thickness is remarkable. Neither, he 

 continues, could the Sigillarias have been Monocotyledons, for 

 their leaf-scars are too narrow. Thus he arrives, by a process of 

 exclusion, at the Vascular Cryptogams, "that is to say the Eerns, 

 for that is evidently the ouly one of the families of this group to 

 which one could refer the Sigillarias." He argues elaborately in 

 support of this view; the Lycopod affinities seem never to have 

 occurred to him at that time. It was no doubt the superficial 

 resemblances between the Eern-stems and Sigillaria which misled 

 him, though the comparison with Lepidodendron seems to us fairly 

 obvious. 



The incomplete second volume, the parts of which began to 

 appear in 1837, is devoted, as far as it goes, to the Lycopods. 

 He gives a most admirable account of the external characters 

 of recent Lycopods and has a good deal to say about the anatomy, 

 which he illustrates by some capital figures. He does not, 

 however, distinguish clearly between the structure of Lycopodium. 

 and that of Selaginella, or Stacliyrjynand.rum as he called it. He 

 notices the interesting fact that in some Lycopodiums the roots 

 have practically the same structure as the stem (p. 24). 



He is at pains to show that the anatomy of Lycopods and 

 Conifers is essentially different. On the other hand, he lays great 

 stress on the resemblance in the cones of the two groups, saying 

 that in Conifers and Cycads the ovules are fixed on the scales 

 exactly like the "capsules" of Lycopodiacese. In describing the 

 two kinds of spores in " Slacliygynandrum " and Isoetes he calls the 

 large spores " veritable seeds " ; he compares the small spores to 

 pollen-grains but declines to discuss their function (p. 33). 



He is very emphatic on the Lycopod affinities of Lepidodendron, 

 saying that the fossils of which that genus is the type merely form 

 a special section of the family Lycopodiacese. " A fortunate and 

 rare circumstance " had given him an opportunity of studying 

 the internal structure. This, of course, refers to the famous 

 Lepidodendron Harcourtii, first described by Witham in 1832, 

 and afterwards by Lindley and Hutton in vol. ii. of their ' British 

 Fossil Flora,' 1833. Brongniart's account of the structure 

 shows a great advance on the previous descriptions. He was the 

 first to recognise the ring of wood, with its smaller elements 

 towards the outside. Oddly enough, this principal feature of 

 the anatomy had been overlooked, or at least not clearly 

 distinguished, by the English writers. The relation of the leaf- 



