72 BOTANICAL GAZETTE [july 



sections and building up the corresponding structures in clay 

 (figs. 16, 17). Vegetative portions were selected for reproduction 

 and the method of study gave some interesting results. The 

 model of the stele of L. varium was almost a complete cylinder, 

 with occasional breaks in the peripheral shell of xylem (fig. 16). 

 This structure is due to the growth of the xylem around the periph- 

 ery. There were some indications that this growth is related 



Fig. 15. — L. varium: transverse section of the strobilus-bearing branch, showing 

 the radial arrangement; px s protoxylem; p, phloem; X300. 



to the passing out of the leaf traces, since often there would be a 

 leaf trace at the point of meeting of two of the peripheral growths. 

 There was too much irregularity in the matter of leaf traces to 

 justify making any definite statements, since leaf traces pass off 

 from the protoxylem points and from the peripheral growths also, 

 with no apparent general relation to the growth of the xylem. The 

 study of the clay models showed branching and fusions of the 

 protoxylem points to be quite common. The relation of xylem 



