398 Dar bishire. — Observations on M ami liar ia elongata . 
the root during the former would tend to fix the plant more firmly in the 
soil, as the lower tips of the roots would of course be firmly attached to 
the particles of soil by the root-hairs. 
This fixing of the plant to the soil has been described by Michaelis for 
Anhalonium fissnratum , Lem. In this case, the whole plant during the dry 
season is bodily drawn into the soil ( 19 , p. 22, PL III, Fig. 12). This 
plant appears also to exhibit annular thickening in the wood-portion of the 
root ( 19 , pp. 17, 18). 
It is not surprising that in the case of the root of Mamillaria we find 
not only the metaxylem but also the protoxylem developing annular to the 
complete exclusion of spiral elements (PI. XXVI, Fig. 32). The annular 
tracheids can probably also store water. 
As we ascend into the stem we find the annular tracheids of the root- 
metaxylem giving way to beautifully developed spiral tracheids. These 
are short and broad, and are not in any way of the type to which the 
protoxylem-tracheids of stem and root in other normal plants belong. In 
the large spiral tracheids of the succulent stem of our plant, I see as much 
water-storing as water-conducting organs. The spiral, again, is a structure 
which allows of an elongation of the cell to which it belongs, but probably 
not of a great subsequent contraction. The continuity of the lignified 
thickening is, as we can see by comparing the wood of the root and shoot, 
not necessary for the conduction of water. But of course the method by 
which the water passes along in the root may possibly differ from that by 
which it passes along in the shoot. 
The libriform cells met with at the point of attachment of root to shoot 
may safely be put down as representing mechanical elements to strengthen 
the firmness of the plant at the point where it is fixed in the soil. For 
this reason we find these cells developed in the later-formed wood-portions 
of the bundle. 
The large parenchymatous cells of the cortex of the main plant-body 
are no doubt cells which store water. I do not think, however, that they 
will exert a very strong osmotic pull on the wood-elements. 
The ground-tissue cells immediately surrounding the bundles are 
flattened (PI. XXVI, Fig. 21, s) and would possess a comparatively small 
vacuole. It is therefore probable that the transpiration stream is drawn 
osmotically to other parts of the plant where the water and its solutes are 
primarily more urgently needed, namely to the tubercles. 
The large central vascular bundles send off branches, which consist, in 
their wood, almost entirely of spiral tracheids, narrow and apparently very 
long. 
These spiral elements lead to the tubercles, and here they separate into 
two systems, namely a cortical and a medullary system — if I may use the 
terms cortical and medullary in this sense. 
