DAWSOIf^ — 6IGIL1AKIA, CALAMITES, AND CALAMODENDKON. 153 



distinguish the structures characteristic of the subgenera of Sigil- 

 laria, or absolutely to separate these from those of certain peculiar 

 conifers on the one hand^ and from those of the higher acrogens on 

 the other. Young and succulent stems of Dadoxylon may have 

 much resembled Sigillaria in their structure. Young stems of 

 Sigillaria proper may have approached closely to those of Favu- 

 laria ; and since I have shown.* that the branches of Favularia 

 resemble Glathraria in their scars, this last may have presented a 

 stiU feebler type of internal organization. Further, there is, as I 

 have already stated, reason to believe that some of the species 

 referred by palseobotanists to the ClatJirai-ia-diyiaion are really 

 forms of Lepidojpliloios. These difficulties, in connexion with the 

 defective state of preservation of specimens, may excuse many differ- 

 ences of opinion, though I think the facts already stated in this paper 

 are sufficient to put all students of the subject on the right track in 

 regard to at least one leading type of these plants, and to remove 

 some of the more fruitful sources of error. 



We may now proceed to inquire what light the structures of 

 Sig'dlaria throw on its affinities. On this question, taken in its 

 - most general aspect, there have, I believe, in modern times been 

 only two opinions, the views as to alliance with EttpJiorbice and 

 Cacti held by some older botanists having been given up. Some 

 botanists, conspicuous among whom is Brongniart, hold that Sigil- 

 larice were gymnospermous plants, allied to Cycadacese. Others are 

 disposed to regard them as acrogens, and as closely related to Lyco- 

 podiaceae. 



In favour of the latter view may be urged the apparent associa- 

 tion with Sigillaria of certain strobiles resembling those of Lepido- 

 pldoios, the points of resemblance between the tissues of Favularia 

 elegans and those of Lepidodcndron, and the resemblance of certain 

 Sigillaricp, or supposed Sigillarice, of the Cluthraria-iy^Q to Lepido- 

 jjhloios. 



In favour of the former view, we may adduce the exogenous 

 structure of the stem oi Sigillaria, and the obvious affinity of its 

 tissues to these of Conifers and Cycads, as well as the constant 

 association with trees of this genus of the evidently phanerogamous 

 fruits known as Trigonocarptim and Cardiocarpmn. On the other 

 hand, the resemblance to Lepidodendron may be shown to depend 

 merely on comparisons of a part of the tissues of Sigillaria witli 

 those of that genus. Grave doubts ma)' also be entertained as to 

 whether strobiles of Lcpidop)hloios, and even stems of that genus have 

 not been improperly mixed up with Sigillaria. 



It is probable that all botanists who have studied these plants, 

 might agree that, if not Gymnospcrms, they at least present points 

 of affinity with them, and might be regarded as in some sense a 

 link connecting them with Acrogens. Supposing this much to be 

 admitted, important questions remain as to their possible relations to 

 the modern Conifers and Cycads. The higher Sigillaria} unquestion- 



* " Conditions of Deposition of Coal," Quart, Journ. Geol. Soc. vol. xxii. 

 p. 130. 



