io WATSON, Two Species of Lepidodendron Harcourtii. 



strands and the course of the latter have not been fully 

 made out, owing partly to the small projection of the 

 protoxylem, and partly to the small number of really 

 good transverse sections. 



It seems, however, that the whole protoxylem mass 

 bent out and became the leaf trace, which is here com- 

 posed of short spiral tracheids. {Fig- 3, Plate I.) 



Certain places in the transverse sections suggest that 

 the protoxylem strands anastonosed occasionally. 



The leaf trace immediately after leaving the wood of 

 the axis is a small bundle of about a dozen tracheids, the 

 smallest being central in position. 



The leaf trace is thus apparently mesarch at this 

 point. {Fig. 6, Plate III.) 



Further out beyond the phloem zone the xylem of 

 the trace has altered in form, being now tangentially 

 elongated, and consisting of tracheids of almost uniform 

 diameter. {Fig. 9. Plate III.) 



No well preserved leaf trace in mid cortex has been 

 met with. 



In the outer cortex the wood of the leaf trace consists 

 of a small circular, and apparently mesarch group of 

 tracheids, which is surrounded by a uniform soft-celled 

 tissue, part of which is probably phloem. To the outer 

 side of the phloem is the group of dark cells which misled 

 Williamson. {Fig. 8, Plate III.) 



This group of dark cells is not well enough preserved in 

 the longitudinal sections to be worth describing. Outside 

 this group of dark elements and separated from it by a 

 layer of small cells is the parichnos. {Fig. 8, Plate III.) 



The xylem of the leaf trace begins to be augmented 

 by transfusion tracheids even before it passes into the 

 periderm, and this group of accessory tracheids is of far 

 greater dimensions than the whole bundle. {Fig. ^Plate II.) 



