346 Groom. — On Thismia Aseroe ( Beccari ) 
nucleus because that substance is present there in the optimum 
proportion in that cell. Miyoshi showed that the attractive 
action of a solution varies with its concentration, there being 
an optimum strength, but the repellent action of a solution 
which attracted when dilute only sets in at high concentrations. 
The question arises as to the source of this chemotropically 
active substance in Thismia. Obviously it is not conducted 
along the hyphae themselves, so it must be absorbed from the 
cell in which the hypha lies. It must therefore be either derived 
from cells lying outside, or be manufactured in the cell in which 
the hypha actually lies. I shall endeavour to show that the 
latter is the case. In the first place, if the substance came from 
cells lying exterior to the mediocortex or even exterior to the 
exocortex, we can conceive of no reason why the hyphae should 
penetrate these deeper tissues at all. On the other hand, we 
should expect to find the cells outside would be most richly 
supplied with hyphae and the deeper cells successively poorer 
in hyphae. The reverse is actually the case; every cell 
(excepting raphide-mucilage cells) from the exocortex to the 
inmost limit of the mediocortex contains hyphae, whereas 
comparatively few cells of the sheath have hyphae. Thus the 
chemotropic action of the cells is weakest in the sheath and 
strongest in the cortex. In the second place, inside a single 
cell the chemotropism towards the nucleus, as judged by the 
curvature of the hyphae towards the nucleus, is weakest in 
the sheath, and strongest in the mediocortex, whilst the 
exocortex forms a layer transitional in this respect. In the 
third place, it would be apparently contrary to the laws of 
osmosis that a liquid absorbed from outside a cell should 
accumulate in greatest concentration in the neighbourhood of 
the nucleus. But this is not a conclusive argument, because 
in plants we often find liquids distributed in a manner 
apparently in opposition to physical laws (e.g. sugar in nerve- 
parenchyma of leaves). 
Thus we are driven to conclude that the chemotropically 
active substance attracts the hyphae , and is manufactured in 
the cell infected, and particularly in the vicinity of the nucleus of 
