io8o 
Thompson.— The Anatomy and 
also show that the two strands are inserted separate!}/. Each arises directly 
above one of the smaller internodal bundles, a piece of the girdle at this 
point simply turning out to form a strand (see Text-fig. i). Sometimes 
a thickening of the girdle runs up from the small bundle and out with the 
trace, indicating that the small bundle as a whole originally formed the 
trace. 
That part of the girdle between the two strands of each trace continues 
up as a definite bundle, so that in transverse sections the two strands are 
separated by a block of wood. This feature is shown in Fig. 5. The same 
bundle is seen in tangential section in Fig. 7. Above the traces it divides, 
and each part unites laterally with the girdle to form the outgoing branch. 
Text-fig. i. E . distachya. Reconstruction of the course of the primary fibro-vascular bundles. 
The bundles of the succeeding internode are organized from the nodal 
wood in a definite manner and position with respect to those of the lower 
internode. On each side of both branch gaps a large bundle is organized. 
The two pairs of large bundles of each internode are thus accounted for. 
Between the pairs of large bundles two small ones are then formed, and the 
two pairs of small ones are then accounted for. The result is the same 
organization as in the lower internode, except that the large bundles are now 
in the position of the lower small ones and vice versa. At the succeeding 
node the story is repeated, except of course that the leaves alternate with 
those below. 
It is obvious that the bundles of the upper internode, although their 
position with respect to the lower ones is definite, cannot be regarded 
as continuations of them. Nor can it be said that any bundle of the inter- 
node passes out definitely as a leaf-trace. It is true that in some specimens 
