PRIMARY STRUCTURE OF CERTAIN PALAEOZOIC STEMS. 349 



for the great majority of the strands ; in a few of the smallest the protoxylem may lie 

 on one side or the other (e.g., the strand 13a shown in phot. 9). 



At the point where a double bundle comes into contact with the secondary wood 

 previous to passing out into the latter, the structure is less regular. Fig. 14 shows 

 this in the case of the double strand numbered 13. The smallest elements of the two 

 xylem-strands are here directed towards each other, and some of them abut directly on 

 a wedge of secondary wood. In a section cut a little further up the stem, where this 

 bundle is shown entering the wood on its outward course, it appears as a single strand, 

 the two half-bundles having re-united. It is possible that their temporary separation 

 may have been due merely to the intrusion of dilated parenchyma. 



The similar outgoing strand between bundles 5 and 6 (l.t. in diagram 5 and phot. 

 5), though obscure, and difficult to distinguish from the adjacent secondary wood, is 

 clearly a single one. It is remarkable how small were the dimensions of the leaf-trace 

 (as we must assume it to have been), at least as regards its primary xylem. Possibly it 

 was supplemented on its outward course by an arc of secondary wood, as was the .case 

 in Poroxylon, * but at present we have no information as to any of the external 

 tissues of our fossil. It will be noticed that behind each of the outgoing strands 1 3 and 

 40 (diagram 5) there is a small xylem-bundle deeply embedded in the pith (12a and 

 39a respectively) ; the strand 5 stands in a similar relation to the leaf-trace, l.t. The 

 strand 12a is shown in detail in fig. 14. In this case there is a second deep-seated 

 strand near by (13a, shown in phot. 9). In the uppermost of the five successive trans- 

 verse sections of the stem which we possess, where the strand 13 is beginning to enter 

 the wood, the two deep-seated strands, 12a and 13a, are approaching each other as if 

 about to fuse, but there is no evidence to show whether such a fusion was of general 

 occurrence. In any case it is natural to regard the deep-seated strand behind the leaf- 

 trace as a reparatory strand, destined to constitute or contribute to the next outgoing 

 bundle of the same orthostichy. 



The phyllotaxis was no doubt a spiral one, and very probably complex, as suggested 

 by the large number of primary strands. 



There is evidence that the primary strands occasionally branched and anastomosed 

 with one another. This is best seen in a tangential section, passing through the outer 

 part of the pith, and touching in places on the secondary wood, represented in phot. 7. 

 Several of the primary xylem-strands are shown ; the double strand, which corresponds 

 almost exactly in size and position with some of those shown in transverse section (cf. 

 fig. 14), is in contact with the innermost tracheides of the secondary wood, and may 

 probably represent a leaf-trace about to pass out. An oblique anastomosis between the 

 strands of this pair is present. 



Another strand appears to be branching, and at several places single tracheides are 

 seen diverging from the xylem-strands in various directions, probably to form a 

 connection with others. 



* Cf. Pitys Withami, below, p. 355. 



