28 STUDIES IN FOSSIL BOTANY 



shown in certain specimens which have retained their' 

 foliage, and we know that each of these whorled leaves 

 actually received a single bundle from the stem (see 

 Fig. 12). Secondly, the outgoing bundles form the 

 direct continuation of those which traverse the inter- 

 node below, precisely as is the case with the leaf-traces 

 of Equisetum. And then again we have the argument 

 of exclusion. If these bundles are not leaf-traces, 

 what are they ? They must in that case be either 

 traces of branches or of roots. But as we shall see 

 below, the connection of these organs with. the stem is 

 now well known, and is quite independent of the bundles 

 in question. By all these considerations the inter- 

 pretation of these outgoing bundles as leaf-traces is 

 now fully 'established. If a section be cut tangenti- 

 ally through the wood, it intersects these bundles 

 transversely (see Figs. 8 and 9). They also appear (cut 

 longitudinally) in the transverse sections of the stem, 

 which pass exactly through a node. A radical section 

 in the plane of the outgoing strand shows that the 

 primary bundle passes out as a whole on its way to 

 the leaf, and can be followed through the secondary 

 wood. The whole course of the leaf-traces is thus 

 made clear from every point of view. 1 



The next point to be considered is the structure of 

 the secondary wood. We find that the wood as a whole 

 is divided up into the bundles, and the principal rays 

 between them. The behaviour of these rays in passing 

 through the wood is a matter of some importance, for 

 great emphasis has been laid on it in discriminating 



1 For further illustrations of all these points, see the memoir by 

 Williamson and Scott, above cited. 



