58a SECONDARF CHANGES. 



cases 6-7, peripheral rings without the central main ring. Radlkofer calls the first- 

 named structure a compound xylem, the second a divided xylem. In both cases each 

 woody ring is surrounded by a normal, permanently active zone of cambium, and 

 also by a normal layer of bast. Each encloses a pith, which is often inconspicuous, 

 immediately around which Jussieu showed that there were spiral vessels. Following 

 them in a longitudinal direction, it is found that the wood- and bast-rings run 

 side by side in the internodes ; at the nodes they are connected one with another, 

 in a manner to be described below, so that as regards the compound xylem the 

 opin-'on might arise that the small outer ones might be the wood and bast of 

 axillary branches joined longitudinally to the main stem. The .composition de- 

 scribed is found in the stems from the commencement of differentiation of tissues 

 onwards. (Fig. 232.) 



We are indebted to Nageli's investigations of the development of a number of 

 forms with compound xylem for a proper understanding of this structure. 



The main results of these are summarised below, partly in the author's own 

 words : we must forego even a short statement of the individual phenomena in one, 

 species, and the differences in different species, because of their complication, 

 reference being made to the original work. 



The stalks or stems of the plants in question are ridged from the very first; 

 the leaves are spirally arranged ; from the axil of the leaves arise a branch and a 

 tendril as axillary shoots. The course and development of the primary bundles 

 correspond to the main rules which hold generally for Dicotyledons (Chap. VIII). 

 At each node three bundles of the trace, and two axillary bundles of the trace, enter 

 the stem. They show generally a tangentially-oblique course, which deviates more 

 or less from that of the angles of the stem. The median bundles have a tendency 

 to unite into three sympodia. Both sorts of bundles of the trace have usually a 

 radially-oblique course; from their point of entry into the node to their insertion 

 on a bundle of a lower node they pass further and further from the surface of the 

 stem. All bundles are collateral, often perhaps^ bi-coUateral. 



In many forms, which correspond closely to those here to be discussed, such as 

 the Cardiospermum, described by Nageli, PauUinia sp., and Serjania Mexicana, the 

 bundles of the trace appear in the transverse section of the young internode at very 

 different distances from the middle ; they surround a prismatic pith, which has 

 projecting angles, and even infolded sides. Nevertheless, they are connected by 

 a normally arranged, general layer of cambium, and together form a simple wood 

 and bast body of the Dicotyledonous type, which is normal and remains so, though 

 of angular and infolded outline. 



In other forms, however, with a strongly oblique course of the bundles, the 

 infoldings of the primary ring of bundles become so deep that single groups of 

 bundles are as it were nipped off, to continue the metaphor, externally from the ring, 

 and excluded. The bundles of such a group are at various distances from one 

 another, and from the middle of the internode. The bundles of the group, which 

 are external as regards the latter, have their xylem and phloem placed normally 



' Compare Nageli, /, c, p. 35. 



