of the Fossil Plants of the genus Sigillaria. 281 



fossil not only approaches to the Cycadeous Gymnosperms in some re- 

 spects, but in others it manifests a leaning to the Vascular Crypto- 

 gams. I am even inclined to grant it a closer resemblance to the latter, 

 than is admitted by Brongniart. The consideration of this point, how 

 ever, must be waived until it is educed in its proper place ; and, for the 

 present, I will content myself with the admission that the genus Sigil- 

 laria comes in between the two groups just named. I presume this is 

 not going beyond what Brongniart will agree to. This admission granted, 

 it follows, that the foliage of this fossil would neither be truly Cyca- 

 deous, nor truly Vascular Cryptogamic, but rather of a character de- 

 parting from the Ferns, and approaching the Cycases. 



Now, whether are we to consider Pecopteris or NenropUrh in this 

 light \ What has already been stated with reference to these fossils, 

 will have somewhat biassed the mind in favour of an answer to this 

 question ; but, in order that we may be further supported, let us exa- 

 mine some other fossils, which, from their affinity to the last-mentioned 

 genus, require our next consideration. According to Brongniart, the 

 genus Neuropteris is characterized by having the leaflets adhering to the 

 rachis only by the median part of their base, and their veins, which are 

 dichotomous, diverging from this part.* 



It is evident, however, if we are to consider such forms as Neuropteris 

 conferta, N. decnrretu, N. elegan?, and some others, as belonging to this 

 genus, the above diagnosis will require some modification. Sternberg has 

 already done this, by placing the foregoing in his third section, which he 

 describes as having the leaflets attached by their entire base, and some- 

 times decurrent.t 



This section appears to form the link which connects Neuropteris with 

 Odontopteris — the genus requiring our next consideration. A glance at 

 Neuropteris Villiersi (Brogn.), Odontopteris crenulata (Brong.), and O. 

 Braardi (Brong.), will shew hoM^ difficult it is to distinguish the one 

 genus from the other. 



We have now arrived at a group of fossils, which, to use the language 

 of Brongniart, ** is completely different from all living Ferns with which 

 we are acquainted." Their leaflets are very thin, and adhere by the 

 whole of their base to the rachis ; they have no, or almost no, midrib ; 

 their veins are equal, simple, or forked, very fine, and most of them 

 spring from the rachis. No known living Fern possesses leaflets and 

 veins of this character.;}: Species of this genus are exceedingly rare ; only 

 seven or eight having been described. Odontopteris Lindleyana (Stemb.)§ 



• " V^getaux Fossiles," tome i. p. 226, and •' Prodrome d'une Histoire des V6- 

 g^taux Fossiles," p. 52. 



1 Essai d'un Tableau Geognostico-Botanique de la Flore du Monde primitif. 

 Par Comte G. Sternberg, 1835, p. 74. — In this section of Neuropteris, several so- 

 called Pecopterides ought, I think, to be included. 



+ " Vegetaux Fossiles," tome i. p. 250. 



§ Odontopteris dubia (Lindley and Hutton), vide " Fossil Flora," pi. 40. 

 VOL. XXXVI. NO. LXXII. APRIL 1844. T 



