154 



PU0CEi;i)IN09 OF TUK OEOLDUICAL SOCIETY. 



ably rcHcniblc Cycads in tho Btructnro of their Htoms. Their long 

 ri^id iiiurow k'livos nisiy ]»o coinpiiivd to siiiglo piiinio of tho loavi's 

 of Cycads. Their cord-liU*' '-ootU'ts, jis 1 have ascertained hy actual 

 comparison, are similar to those of (lycads. If tlu-ir fruit was of the 

 nature of Cdrdlocarjnim or Trit/oiuicarpxm, this uIbo would corre- 

 spuiid. They differed prineipally in the division of the stem holow 

 into those remarkable underjiround branches, tlie Sfli/tn<iri<i', and in 

 tlie ^reat upward extension and, in some instances at least, raniiti- 

 cation of the stem. The former may be rej^arded as a special modi- 

 fication connected with their i)ecidiar habitat. The latter may be 

 interpreted as a modification either tending backward to the Lycopo- 

 diaceie or forward to the Couiferic. Since, so far as we at prcsonl 

 know, tho ramification prevails clp'efiy in tho lowc^r forms, the 

 foruKM' may be the more correct view. It is (;ven possible that the 

 iSiyilhti'Uf may include forms bridf^infj: over the space between the 

 hiji;hcr Acrogcns and the Uymnosperms. Viewed in this way, the 

 typical ribbed Si(/HI(iri(i' point downwards throufijh Calamodeailron 

 and Cahnnitesto the K(piisetaceic, and the Favuldrid- and Cldfht'drla- 

 types point throufih fjCi>i(((ij)hlolosi\m\ L<i>hlQih'ti<lron to Lyco{)odiacea?. 

 In tho upward direction their atfinities j)oint both towards Conifers 

 and Cycads. As our kjiowh'dge of the structure of indiWdual species 

 of S'ltjUland increases, we may hope more certainly to trace the links 

 of these affinities. It is, however, to bo observed here, by way of 

 caution, (1 ) that, of the ])lants reckoned among the several genera or 

 subgenera oi Si(/ill(iri<t', some may eventually prove to Ite gymno- 

 spermouH and some cryptogamous, and {'2) that, as wo shall find in 

 the next grcmp to have been actually the case, som(! of these i)lant.s 

 may, with a cryptogamous fructification, have i)rcsented a structure 

 of stem more compkjx than that found in modern pkints of similar 

 grade. 



2. Calamodkndron and Calamites. 



Calamitcs are among the most abundant fossils of the Carboni- 

 ferous period, and occur also in tho Devoidan ; and from their pecu- 

 har habitat and mode of growth, they arc not only jjroserved as 

 flattened steins, but also occur in immense numbers standing on the 

 beds on which they grew. 



They have naturally been regarded from the first as allied to 

 IViuisetacea) ; and this opinion is ably and, indeed, conclusively 

 maintained by Schimper in liis recent work*, and has been illus- 

 T^^-ffi 1 -^ ^J^ I't^ccnt description of tlic fruit by Mr. Carruthers. 

 Uithcuities have, however, arisen from the fact that some stems 

 rcr-rded as Calamites have been found to bo surrounded by a thick 

 woody cylinder composed of discigerous and pseudo-scalariforra 

 tissue similar to that of tlie type of i^iylUarui above described, 

 borne botanists have regarded these last m distinct from the time 



r!lZ T T ^r' 1'^'.''^ ^^"" "^ the genus Calamitea, Cotta, or 

 Calamodenclron, Brongniart ; and Williamson has recently proposed 



* Palcontologie Vcgetalc. * 



