( 196 ) 
branches is less than that in the bark. The quantitative determination., 
! wh,ch took P lace ^ ‘he same way as in Vaccinium, were carried 
out on a single tree of the variety Bonne Louise d’Avranches. 
The glucoside is formed in the leaves, and the formation begins 
in the young shoots as soon as they themselves begin to assimilate. 
On comparison of the amounts in the halves of a leaf in the 
morning and in the evening, the glucoside is found to increase during 
the day and to diminish during the night. This diminution during 
the night is again accompanied by an increase of hydroquinone and 
the amounts are in the proportion of 100; 18 % whereas the molecular 
weights are in the proportion of 100 : 40. A portion of the hydro- 
quinone, which has been liberated in the nocturnal hydrolysis thus 
appears to be directly transformed in metabolism, or else we must 
assume, that a part of the arbutin is transported as such. 
During the summer the amount of arbutin increases in the bark, 
from 1.5 °/o to 4.5%, so that we may safely infer, that the arbutin 
is deposited there in order to be used up in the formation of new 
shoots in the spring. At that season the amount in the bark diminishes 
greatly (about 70% of the total) and increases less markedly in the 
young shoots, so that the total diminution from March 30 th to May 2 nd 
amounted to about 27 %• During this process the amount of free 
hydroquinone is greatly increased by fermentative arbutin hydrolysis 
and from April 26 th to May 2 nd , for instance, the decrease of arbutin 
was to the increase of hydroquinone in the proportion of 100:40, 
exactly that of the molecular weights; in other words the whole 
of the aromatic product remains localized in the cells*)- Afterwards, 
when assimilation has begun in May, hydroquinone is rapidly worked 
up and arbutin is formed at its expense. Hence we find Pfeffer’s 
hypothesis once more confirmed: hydroquinone combines with the 
transport substance glucose, with formation of the reserve material 
arbutin. 
In the case of Sahx purpurea I found in 1902 that by slow 
dessication 25 •/, of the salicin disappeared from the bark. This 
pointed to the presence of a salicin-splitting enzyme, but at the time 
I was not able to demonstrate it. On continuing the experiments 
9 In Salix purpurea also the relation of salicin decrease to catechol increase 
in day and night experiments differed more from the theoretical than in the case 
of the opening of buds of branches placed in water. 
2 ) Most of the hydroquinone formed, was in the young shoots; this might 
therefore be interpreted as an indication, that a transport of arbutin takes place in 
this case. 
