16 PASSAGE OF FOOD FROM LEAVES TO STEM. 
case of the box, so with the pine, rhododendron, and laurel, 
_the current in these vessels was found to be directed downwards 
towards developing buds. But as the experiment was made on 
the upper and younger portions of stems, I cannot say whether 
this would also happen in older portions. 
I have treated of the stems first, chiefly because there was more 
uniformity in the course followed by the eosine in them than in 
the leaves. The extent of development of the different tissues 
being greater makes it easier to experiment on stems, and also to 
observe the results. But as one would naturally, if he were 
trying this experiment on a growing plant, think of injecting into 
the leaves, I shall state what happened in the cases I tried. At 
first I inserted the eosine into the green cells, with the result that 
the cells surrounding the puncture died, and showed as brown 
spots a quarter of an inch in diameter. The effect produced in a 
hyacinth leaf was equally disappointing. The eosine was injected 
one and a half inches from the tip and a quarter of an inch from 
the margin. The neighbouring cells turned flaccid, but not 
brown, and no cork was formed as in the case of evergreen leaves. 
The remainder of the leaf was in no way affected, and as the leaf 
continued to grow it became contorted at the tip. The eosine 
has evidently an injurious effect when introduced into cells in 
which assimilation is going on, but when introduced into the 
under side of the midrib it has no such bad effect. It is not easy 
determining how far to insert the needle, but leaves are more 
plentiful than stems, and are easier to section, and you can always 
see by examining a section cut at the point of insertion how far 
the needle has entered. If the eosine enters the sub-epidermal 
tissue only there is no result. The laurel leaf is an interesting 
one to experiment on, on account of the living cells which extend 
in a radial direction between the xylem vessels, so that they are 
in connection with both the phloem and the xylem. The eosine 
inserted into the phloem travels downwards very slowly, compared 
with what it does in the sclerenchymatous tissue outside of the 
phlcem, or in the living cells between the vessels. In the leaves 
the eosine which enters the vessels from the parenchymatous 
cells travels upwards. In the other cells the direction is a down- 
ward one. When conducting the experiments I did not take 
notes on the rate of travelling, because it was to be expected that 
