THE PREPARATION AND CHEMICAL COMPOSITION OF TOFU. 213 
pressed, and then cut into tabular shape. (I) I have tried to 
arrive at a satisfactory explanation of the nature of tofu, and have 
found that the salt-brine does not act by its chloride of 
sodium, but by the calcium and magnesium salts which are 
in it ; for we can at once obtain the precipitate from the 
milky liquid if we add a little calcium nitrate or magnesium 
sulphate, (2) while we can not obtain any separation or precipita¬ 
tion by adding even considerable quantities of sodium chloride 
or sodium sulphate. (3) However, ammonium sulphate will, as 
with cow’s milk, bring on precipitation, if we add 60 % of it to 
the milky liquid. This precipitate is so voluminous that the 
liquid seems to solidify. The peculiar state of the solution of 
soya bean casein resembles that of milk ; thus, if milk is 
dropped upon porous clay, the liquid will be absorbed by the 
clay, but the casein and fat will remain on the surface, and the 
same is observed with the milk from soya beans. (4) 
I have analysed a sample of the salt brine used for tofu mak¬ 
ing and found it to contain, besides chloride of sodium, 27.9 % of 
chloride of magnesium and 7.0 % of chloride of calcium. As 
casein has decidedly an acid character and is only very little 
soluble in water in the free state, it seemed most probable that 
the aqueous extract of soya beans contains a sodium or potas¬ 
sium compound of casein (5) yielding insoluble calcium and 
magnesium compounds, which constitute tofu. 
(1) In this way about one fourth of the total amount of proteid in soya beans is 
obtained in the tofu. 
(2) An excess of magnesium sulphate is to be avoided, as it would redissolve the 
precipitate. 
(3) If the liquid is warm and saturated with sodium sulphate, a very slow and 
imperfect separation may be noticed. 
(4) In order to see whether a product similar to Swiss cheese could be obtained 
from the crude soya casein or tofu , I infected 50 grm. of fresh tofu with a small dose 
of pulverised Swiss cheese, and added ten per cent of common salt to the mixture, 
pressed it in cloth, and allowed it to stand in a moist beaker glass for several 
months. The product resembled, only to a limited extent, the cheese from milk, but 
further experiments with addition of small quantities of milk sugar are intended. 
(5) I observed that soya bean casein is easily digested by pepsin solution 
acidulated with 0.2 o/ 0 hydrochloric acid, without leaving any insoluble residue; 
in this it resembles human casein, which, according to the valuable investigations 
of Wroblewski (Berne, 1894) yields, unlike cow’s-casein, no paranuclein. I found 
also that peas and hôrse-beans yield aqueous extracts of similar behaviour to that 
of soya beans. 
