642 MR JOHN M'LEAN THOMPSON ON 



arise de novo* and it will be shown immediately that, in the liberating of the 

 leaf-traces of Platyzoma, endodermis is formed where and when required. It is then 

 not impossible that the inner endodermis of Platyzoma is of relatively recent origin, 

 and has been at no point in descent in connection with the outer endodermis, either 

 through leaf-gaps or perforations. If this be so, the pith of Platyzoma cannot have 

 been of cortical origin, and the stele cannot be interpreted as a reduced solenostele. 



We come now to the origin and nature of the leaf- traces. There are both large 

 and small leaf-traces (tigs. 1, 4, and 13). The largest traces arise typically from the 

 upper surface of the stele, but occasionally their points of origin are at a slightly 

 lower level. The smallest traces originate from the lower surface of the stele, and 

 between the largest and smallest types traces of intermediate size exist. These 

 latter are inserted typically upon the sides of the stele. There is, then, a gradation 

 in the sizes of the leaf-traces as one passes from the upper to the lower surface of the 

 stele. Slight differences of detail occur from trace to trace, but one general method 

 is adopted throughout in the formation of the leaf-traces. It will suffice to describe, 

 in detail, the departure of a large leaf-trace, and to indicate briefly such points as are 

 of interest in the smaller types. 



The first indication of the formation of a leaf-trace is the marking out and pro- 

 jection of a peripheral, tangential band of tracheides from the outer xylem-cylinder 

 (fig. 41). Beneath this band of tracheides a definite mound of parenchyma has 

 formed, while above it the phloem, pericycle, and endodermis have been raised in 

 a gentle curve. As the band of xylem becomes more distinctly curved and projecting, 

 the band of phloem, displayed upon its convex side, becomes thinner. Very soon 

 the xylem band begins to free itself from the xylem of the stele, and to pass out as 

 the crescent-shaped xylem of a leaf-trace (fig. 42). It may sever its connections 

 with the stelar xylem simultaneously at both sides, or it may become free on one 

 side more rapidly than on the other. But in any case, while the phloem which clothes 

 the outer surface of the leaf-trace xylem has become thinner, no diminution of 

 phloem has taken place at the points where the leaf-trace xylem has freed itself. 

 In fact, no sooner has an edge of the leaf-trace xylem become free, than phloem 

 begins to appear in the gap which separates it from the stelar xylem, and, as the gap 

 widens, a sheet of phloem appears in the parenchymatous mass lying beneath the 

 leaf-trace xylem (compare figs. 42 and 43). This sheet of phloem is an inward con- 

 tinuation of the stelar phloem beneath the developing leaf-trace. It is destined to 

 meet a similar phloem sheet projecting inwards from the gap which separates the 

 leaf-trace xylem from the stele at the other edge ; and when these phloem sheets 

 have met and fused, a continuous layer of phloem will be spread once more over the 

 surface of the stelar xylem. Two stages in this process are represented in figs. 43 

 and 44. Meantime the leaf-trace xylem and its accompanying crescent of phloem 

 have continued their outward course, and very soon the phloem of the leaf-trace 



* Ann, Bot., 1913, pp. 236-239. 



