FRANCIS ON CASSAVA. 
119 
same portion, acidifying with nitric acid, and collecting and 
weighing the silver cyanide produced. Fairly concordant 
results were thus obtained, as shown by the following aver- 
age examples. 
I. Prussic Acid from Sweet Cassava. 
Per cent. 
Grains of 
of H C N. 
H C N per lb. 
Volumetric method 
'01992 
1-3944 
Gravimetric „ 
'01924 
1-3468 
II. Prussic Acid from 
Bitter Cassava . 
Per cent. 
Grains of 
of H C N. 
H C N per lb. 
Volumetric method ... 
•02090 
1-4630 
Gravimetric „ 
'01996 
1-3972 
the higher results furnished by volumetry doubtless showing 
the usual error arising from the slight excess of standard 
solution required to indicate the completion of the process. 
The silver solution used was nominally decinormal, but its 
exact strength was fixed with care. 
During the grating of the cassava, which occupied about 
five minutes, a slight loss of prussic acid was made evident 
by its odour. Nevertheless, this was certainly too small to 
seriously affect the correctness of the determinations, since 
the quantity of prussic acid found in the normal juice seldom 
exceeds one part in two thousand of liquid, and such dilute 
solutions bear moderate exposure without being sensibly 
altered. Estimations, made at intervals, of the prussic acid 
in the diluted juice, showed that it suffered no apparent loss 
during two hours while standing in a loosely-covered beaker ; 
but a loss of about one-tenth was found after sixteen hours. 
When fermentation was in progress, the prussic acid soon 
diminished, and in six days became reduced to one-fifth of 
the original quantity. 
