THE ANATOMY AND AFFINITY OF PLATYZOMA MICROPHYLLUM. 643 



becomes free from the phloem of the stele (compare figs. 44 and 45). The endo- 

 dermis and pericycle have moved outwards in order to accommodate the xylem and 

 phloem of the leaf-trace, and when the limit of this accommodation has been reached 

 the leaf-trace is a crescent of solid xylem, with phloem on its convex surface and 

 around its margin (fig. 45). 



When any gap which had arisen in the peripheral cylinder of stelar xylem has 

 been repaired, and when the phloem covers the stele once more as a continuous 

 sheet, endodermal cells appear on both sides of the leaf- trace (fig. 45). These new 

 endodermal cells spread obliquely inwards from points upon the outer endodermis, 

 and, having passed rapidly inwards, unite to form a continuous endodermis which 

 completely shuts off the leaf-trace from the stele (fig. 46). But further, no sooner 

 has this endodermal formation been completed, than the curved portion of the 

 endodermis, which has accompanied the leaf-trace on its outward course, becomes 

 free as a crescent-shaped band, following the general outline of the leaf-trace. The 

 endodermis of the leaf- trace is then, at this point, of an imperfect character, being 

 maintained upon the outer side of the trace, but absent from the inner side. This 

 condition is not, however, long maintained, for an inner band of endodermis soon 

 appears, repairing the gap (fig. 46) and determining the contour of the leaf-trace as 

 it passes out through the cortex. It is here a solid though narrow band of xylem, 

 with a thin layer of phloem upon its convex surface, and a well-developed pericycle 

 (fig. 47). It is well to note that in this process of leaf-trace formation the inner 

 endodermis is in no way involved, and, further, that there exists no point where the 

 stele is not entirely enclosed in outer endodermis. Not only has it been established 

 that the cortex and pith are connected at no point, but it will now become evident 

 that no connection exists between the cortex and the pericycle at the point of final 

 liberation of the leaf-trace into the cortex. The possibility of such a connection has 

 been removed by the local doubling of the outer endodermis as described above. It 

 frequently occurs, however, that when the liberation of a leaf-trace imposes a large 

 drain on the tracheides of the .outer stelar cylinder, a parenchymatous bridge appears 

 passing athwart the xylem-ring (fig. 63). But many chains of parenchyma traverse 

 the xylem from side to side — though their courses may be irregular and narrow, — 

 and in no case do these bridges disturb the inner or outer endodermis. It happens 

 occasionally that the xylem of the leaf-trace is, at the point of departure of the trace, 

 a compact mass of tracheides which appears almost circular in section ; but as the 

 trace is followed outwards, the xylem becomes extended laterally, and when sectioned 

 shows the typical crescent arrangement described above. 



The smallest leaf-traces may now be considered. Two stages in the liberation of 

 a small trace from the lower surface of the cylinder are represented in figs. 48 and 

 49. It will be evident that they reproduce, in miniature, the method of trace- 

 departure described above for the largest leaves. As they pass outwards into the 

 cortex the steps described for the large leaf-traces repeat themselves, and when the 



