654 MR R. KIDSTON AND MR D. T. GWYNNE-VAUGHAN ON 



As regards the peripheral tissues of the stele, there is a stout " xylem sheath" of 

 some six to eight layers of parenchyma (figs. 6 and 8, xy. sh.). This is succeeded by a 

 wide belt of phloem consisting of four to six layers of large and conspicuous sieve-tubes 

 without any intervening phloem parenchyma (figs. 6 and 8, ph.). On the outside of the 

 phloem the tissues are less well preserved, but they are sufficiently clear to demonstrate 

 the complete absence both of a protophloem and of the porose layers that characterise 

 the recent Osmundacese. The absence of the protophloem is rendered all the more 

 sure since this tissue is distinctly recognisable in the free leaf-traces, as will be shown 

 later on. The pericycle abuts directly on the large sieve-tubes, and is about five layers 

 thick (figs. 6 and 8, per.). It is well delimited from the cortex by a dark band of 

 broken-up cells, which no doubt represent the remains of the lignified walls of the 

 endodermis. It probably consisted of a single layer of cells, although at some points 

 two layers of cells seem to have been endodermal in character (figs. 6 and 8, en.). 



The inner cortex consists of ordinary thin-walled parenchyma, and is very poorly 

 preserved. The outer cortex, on the other hand, is heavily sclerotic, and in transverse 

 section its elements appear as very thick- walled cells with rounded lumens (fig. 14). 

 They are not, however, fibrous, but in longitudinal section are seen to be short and oblong 

 with bluntly rounded ends not more than four or five times as long as broad (fig. 15). 



The Departure of the Leaf-trace. 



The leaf-trace departs in a typically protostelic manner, precisely as that of 



Zalesskya gracilis* It is therefore unnecessary for us to redescribe the process here, 



the departure being sufficiently illustrated by the figures 16 and 17 on PI. III. 



While still within the endodermis of the stem the xylem strand of the leaf-trace is 



shortly oblong with rounded ends, and it contains a single deeply immersed almost 



central strand of protoxylem (fig. 18). It retains the same form for some time even 



after the leaf- trace has become free from the stem stele (fig. 19), and then it gradually 



becomes more or less elliptic as it passes outwards. At about this stage some of the 



centripetal tracheae in front of the protoxylem strand on the adaxial side cease to occur, 



and are replaced by thin- walled parenchymatous cells (fig. 20). To put it in another 



way, these tracheae are not succeeded above by other tracheae but by parenchyma. In 



fact, the xylem stops short at this point, and as a result a small isolated island of 



parenchyma is now found on the adaxial side some little distance in front of the actual 



protoxylem group (fig. 21). As the leaf-trace passes out this island gradually enlarges 



(fig. 22) until at length even the most adaxial tracheae stop short and the island 



becomes converted into a bay open on the adaxial side (fig. 23). The disappearing 



centripetal tracheae do not all stop short at the same time, but some are continued 



upwards beyond the rest as isolated elements in the parenchyma of the bay (fig. 24). 



This bay then gradually becomes more and more open, and at the same time the leaf- 



* Part II., p. 225. 



