Vegetative multiplication in Crotalaria burhia 231 
shoot and the normal branches and those regions gain in thickness 
as the accessory bundles increase in number and grow older. The 
reproductive branches, which appear later in these lower regions of 
the main shoot and normal branches, arise directly from accessory 
bundles and have no connection with the central cylinder of the shoots 
on which they are borne, nor consequently with the main strands of 
the root (Text-fig. 1). The same remarks apply to secondary repro¬ 
ductive branches produced on the primary reproductive branches. 
Thus all these reproductive branches differ in origin from normal 
branches. But normal secondary branches which are also produced 
on these reproductive branches are directly connected to the central 
cylinder of the branches on which they arise. So in old plants, the 
original vegetative shoot and the normal branches do not exist, and 
all the shoots are connected only with the accessory bundle system. 
The central cylinder of the main root dies back from above, after the 
original shoot and normal branches are dead, and in old plants, a 
hollow tube of the accessory strands, with decayed matter inside, 
may be seen extending some distance below the ground. In old plants 
the accessory strands also separate into groups owing to the forma¬ 
tion of cork, but they remain connected below. The groups of accessory 
strands, as has been seen, terminate in “ reproductive ” branches, and 
the latter are therefore carried further apart by the separation of the 
groups. The accessory system remains attached to the main root 
below, where the formation of cork has not taken place, and where 
the accessory bundles decrease in number and become less con¬ 
spicuous, until, traced downwards, they disappear. 
Much light was thrown on many points in this investigation by 
the preparation of a skeleton of the root, along with the lower portions 
of the main shoot and branches, by maceration. The skeleton was 
prepared by allowing it to decay in water for a fortnight and then 
keeping it in dilute nitric acid for a day. It was quite clear then that 
the reproductive branches arose directly from the accessory bundles 
and had no connection with the main central cylinder. 
The accessory system in the root at first appeared to be entirely 
separate from the central cylinder, and it was thus difficult to see 
how it could function as a conducting system as it did not seem to 
develop in relation to any absorbtive tissue. It was found, however, 
from sections cut at points where lateral roots originated, that the 
conducting tissues of the lateral roots, though formed primarily as 
a branch of the central cylinder, were connected by a few strands 
with the accessory bundles. The central cylinder becomes functionless 
