'!?84 Contributions towards Establishing the General Character 



our consideration the principal question, as to whether Pecopteris or 

 Neuropteris constituted the foliage of Sigillaria. 



When detailing the observations of Brongniart on the internal struct 

 ture of Sigillaria elegans, it was stated, that the ligneous cylinder, being 

 formed of scalariform tissue, allied this fossil to the Vascular Cryptogams, 

 and that the form of this cylinder, and the radial arrangement of its 

 tissue, approximated it to the Cycadeous Gymnosperms ; the affinities 

 thus clearly indicated, it is obvious, ought to influence us in our re- 

 searches as to what kind of leaves constituted the foliage of this genus. 

 Accordingly, to arrive at a solution of this question, no other method 

 appeared so legitimate as to ascertain, in the first place, which of the 

 coal-measure leaves afforded the most evidence of their correlation with 

 Sigillaria ; this point gained, the next was to see if the remains thus fixed 

 upon indicated a Cycadeous tendency. 



This is precisely the method that has been adopted. Through various 

 considerations, it was made out, in the first instance, that the weight of 

 evidence was in favour of Pecopteris or Neuropteris : in the next place, 

 and in order to ascertain which of these two forms displayed the strongest 

 tendency to depart from the Ferns and run into the Cycases, two inde- 

 pendent modes of argument were pursued ; the one as founded on the 

 essential characters of either Pecopteris or Neuropteris, graduating into 

 those of a Cycadeous leaf; and the other, as founded on the subsequent 

 appearance of Cycadidasy to whichever of these genera might be decided 

 upon. By these means, it was ascertained that Neuropteris possessed 

 the strongest claims in its favour; in fact, there was traced an un- 

 broken chain of affinity from this Fern to a decidedly Cycadeous type, and, 

 at the same time, the required appearance of the latter, subsequently to 

 the former, was made as conclusive as available data would admit. Thus 

 the evidence is in favour of Neuropteris having constituted the foliage of 

 Sigillaria, 



Before concluding this part of our subject, it will be necessary to ob- 

 serve, that, although the previous reasoning has led to the above conclu- 

 sion, it must not be understood that all the fossils usually considered 

 Neuropterides were thus correlated. The genus, as at present constituted 

 — even confining ourselves to those of the carboniferous epoch — may em- 

 brace species belonging to widely different plants : it is probable some 

 may have belonged to Tree Ferns ; others to herbaceous forms of the 

 same class ; while some may have belonged to plants which cannot be 

 included with the latter, — for instance, Sigillaria. Thus, Neuropteris obo- 

 vata (Stemb,), fragments of which occur in immense numbers in the coal- 

 deposits of Mireschau, and Swina in Bohemia, are found associated with 

 what appear to be specimens of a compound rachis of a gigantic frond, 

 having, in some instances, a length of eight feet.* Drifting might certainly 



.♦ Sternberg, in •* Flore du Monde Primitif," parts 5 and 6, p. 74. The divisions of 

 this rachis are alternate, sometimes forming an acute angle with the axis, and occa- , 



