894.] On the Fossil Plants of the Coal-Measures. 123 



In the phloem of this species large elements, closely resembling 

 sieve tubes, are found. 



The larger specimens, which have lost their primary cortex owing 

 to the formation of periderm, have a very root-like transverse 

 section. Some authors have therefore denied that they belong to 

 Sphenophyllnm, and have supposed that they are roots of some 

 unknown plant. This is a mistake, for the large specimens have 

 essentially the same structure as the smaller ones, which still retain 

 the characteristic cortex and leaves of a Sphenophyllum. The inter- 

 mediate conditions are also known. 



Leaves had not been discovered in this species when the former 

 account of its structure was given, in the earlier memoirs. 



The recent researches of M. Zeiller have proved that the fructi- 

 jcation previously described (Williamson, " Organisation, &c., 

 Part XVIII," ' Phil. Trans.,' 1891) as Bowmanites Dawsoni, is that 

 of a Sphenophyllum. In his specimens, strobili, agreeing in all 

 respects with those of Bowmanites, are borne on the stems of the 

 well known Sphenophyllum cuneifolium, Stern b. The fructification in 

 question must therefore be transferred to the genus Sphenophyllum, 

 and is here described under the name of Sphenophyllum Dawsoni. 



The strobilus consists of an axis bearing numerous whorls of bracts, 

 which are coherent for some distance from their base. The very long 

 sporangiophores arise from the upper surface of the bracts, near their 

 insertion, two sporangiophores corresponding to each bract. At the 

 end of each sporangiophore a single sporangium is borne, which 

 hangs down, parallel to the pedicel, somewhat resembling an ana- 

 tropous ovule in position. 



The axis of the strobilus is traversed by a triarch or hexarch 

 vascular cylinder, essentially similar to that of the vegetative stem 

 of Sphenophyllum. At every node vascular bundles are given oif to 

 the bracts. Each of these bundles, on entering the verticil of bracts, 

 subdivides into three. The dorsal branch passes straight out into one 

 the free bracts. The two ventral branches of the bundle supply 



3 two sporangiophores corresponding to the bract in question. 



The bundle of the sporangiophore extends through its whole length, 



coming thicker towards the apex, where it terminates at the base 

 the sporangium itself. 



The cells of the sporangial wall are of great size near the base, and 

 very narrow at its opposite end, where dehiscence probably took 

 ce. 



The spores are numerous in each sporangium, and are all of the 



ime kind. There is at present no conclusive evidence for the exist- 



ice of a heterosporous Sphenophyllum. 



The morphological nature of the sporangiophore cannot be deter- 

 ad with certainty. The various possible views are stated in the 



