STRUCTURE OF 8TIGMARIA. 437 



Should the conclusion of M. Brongniart hold good, and the vascular 

 system of the axis c give origin to the bundles b, then will there be 

 some botanical evidence of Stigmaria constituting a part of Sigillaria. 

 If, on the other hand, it does not prove correct, still its failure would 

 afford neither proof nor even presumptive evidence against these genera 

 holding the respective place of root and stem. 



As, however, I have elsewhere stated, I consider Mr. Bhmey's dis- 

 coveries to be conclusive upon the origin of Stigmaria. That gentleman 

 had the kindness himself to conduct me to the spot where the great fossil 

 trees were, and we cleared the soil from the roots of several well-marked 

 Sigillaria, which roots were indisputably the plant called Stigmaria. 



M. Goeppert, who holds this genus to be an individual plant, generic- 

 ally distinct from any other, refers it to the Lycopodiacece as its nearest 

 allies, but suggests the probability of its being intermediate between 

 that order and the Cycadece. This exact coincidence with the opinion 

 of M. Brongniart respecting the affinities of Sigillaria is sufficiently 

 striking, especially since these authors entertain different views of the 

 nature of Stigmaria. I have, under my general remarks on Sigillaria, 

 expressed doubts of the correctness of this opinion, in reference to the 

 connexion of two such widely different families by so slender and falla- 

 cious a character as the arrangement of the tissues in the stem. The 

 points by which Sigillaria (and Stigmaria) is allied to Lycopodiacea, 

 especially through the Lepidodendra, are probably quite sufficient ; but 

 when we consider that the Cycadea have sexes, a complicated repro- 

 ductive system, and a perfect seed with embryo, radicles, albumen, &c, 

 whilst Lycopodiacea are asexual, have no reproductive system of male 

 and female flowers, and increase by spores, it will be evident that it is 

 by modifications of these important organs that the intermediate family 

 would be indicated. The noble Tree-fern departs from its humble 

 congeners in assuming the aspect of a Palm, and the Cephalotus (or 

 Pitcher-plant of New Holland) has the leaf of Nepenthes; but the 

 Tree-fern does not unite the Ferns with the Palms, and the Cephalotus 

 is not the less a rosaceous plant because it has the pitcher of Nepenthes ; 

 and so we may find that the Stigmaria is no less a fern from its tissues 

 being modified in arrangement. A still better illustration is furnished 

 by some of the natural order Magnoliacea, that have the woody tissue 

 closely resembling Conifera, and the parasitical Myzodendrons of Cape 

 Horn, whose wood is like that of the ferns in many respects. These 

 are genera certainly departing from their immediate allies in these cha- 

 racters, but not therefore allied to the plants which in those particulars 

 they resemble. 



The symmetrical quincuncial arrangement of the scars has been 

 considered by very many observers to be all but conclusive against the 



2 h 2 



