1 2 Marshall Ward and Dimlop. 
glucose-reaction, hence the ferment was not contained in the 
endosperm. 
On cutting sections it was noticed that there was a large 
quantity of a yellow substance in the cells, which however 
differed considerably from the glucoside of the pericarp in its 
micro-chemical reactions. In order to see if it was different 
from the glucoside the embryos of a number of seeds were 
digested as usual, and to this solution was added a solution of 
the testa ; on allowing to stand for twenty-four hours no 
precipitate was obtained, and it was therefore concluded that 
the yellow substance of the embryo was not the same as the 
glucoside of the pericarp. 
Thus in the pericarp there is contained a glucoside, which 
is split up by a ferment contained somewhere in or near the 
testa and embryo into glucose and a crystalline substance. 
It seemed to be proved from the foregoing observations 
that the ferment which decomposes the glucoside (xantho- 
rhamnin), contained in the cells of the pericarp, is localised in 
the testa of the seed ; whether any traces existed in other 
parts was not proved. On adding water to the intact fruits 
the soluble ferment passes out and acts on the dissolved 
glucoside from the pericarp, breaking it up into rhamnin and 
glucose. If this occurs the precipitate will be withheld chiefly 
inside the fruits, thus explaining why the dyers should crush 
their fruits, since it is the insoluble semi-crystalline precipi- 
tate which they want. 
Before proceeding to show where the ferment really is — in 
the raphe, a discovery which I only made some time after — it 
should be stated that the above observations were unavoidably 
put aside owing to the pressure of new duties. I had, how- 
ever, made some observations which led to the suspicion that 
the ferment is even more localised than it had so far been 
shown to be. 
Series I. 
The following experiments were made with ‘ Persian berries’ 
— the fruits of Rhamnus infectorins— obtained from Kew. 
