CARBONIFEROUS LTCOPODS AND SPHENOPHYLLS. 49 



foliage is still attached, where the leaves on the stems are always 

 much larger than those on the twigs — the general rule being 

 the larger the stem the larger the leaves, up to the period at 

 which the leaves are shed. 



In some species the leaves remain attached to the stem for a 

 greater length of time than in other species. In Lepidodendron 

 li/copodioides, Sternb., the leaf-base seems to have remained on the 

 cushion for some time after its upper portion had been separated 

 or decayed, and in this species one can seldom observe the form 

 of the leaf-scar which is placed towards the upper end of the 

 cushion. 



A specimen from Oaks Colliery, Barnsley, collected by Mr. 

 Hemingway, shows a leaf-cushion of this species with a leaf 

 attached (No. 2165). In this state no line of demarcation can 

 be detected between the leaf and the cushion. The upper 

 portions of the other leaves on this example are broken off 

 (probably, in the case of the specimen under discussion, by 

 mechanical means), but the lower part remains attached to the 

 cushion and extends down each side some distance past the apex, 

 giving the upper part of the cushion the appearance as if winged. 

 The lower portion of the leaf is here much wider than the leaf- 

 scar left after the removal of the leaf, as shown by specimens from 

 which the leaf has been shed, from which one is led to presume 

 that it was partially attached to the cushion. In other species 

 the leaf seems to be cut off by some special provision, though in 

 Lepidodendron lycopodioides it seems to wither away. 



In Lepidodendron Ilaidingeri, Ett. (No. 1022), where also the 

 leaf-scar is situated toward the upper end of the cushion, there 

 appears to be a somewhat similar mode of attachment of the 

 leaf. 



The leaf-scar and cushion increase in size with the increase of 

 the stem in growth, though in a few species, as in Le})idodendron 

 Veltheimianum, Sternb. — to which reference has already been 

 made in this connection — the stem increases in girth quicker 

 than the bark, which becomes longitudinally cleft. 



The leaves are often found separated from their parent stems, 

 and as in this condition they can seldom be identified with the 

 species to which they belong, they are placed in a sejDarate genus 

 — the genus Lepidophyllum, Brongt. 



