n 
RESEARCHES ON THE MANUFACTURE, 
fungus by destruction of organic material is very considerable. 
Similar observations were also made by Atkinson who found 
that the difference between the temperature of the air outside 
and inside the cellar amounted to 23-25°C. in December and 
to 5-7°C. even in May, and that the rice on the trays was on 
the 3rd day of the manufacture about 13 °C. warmer than 
the air in the cellar. 
The samples of steamed grain sown with spores were 
made air-dry in a water-stove, proper measures being 
taken to prevent by a sudden exposition to boiling heat 
any development of the fungus. In the koji we determined 
the total nitrogen, albuminoid nitrogen, ammonia, free acids, 
alcohol, and the solubility in small samples of the fresh sub¬ 
stance, rubbing weighed portions with some pumice-stone in 
all cases in which the koji had to be extracted. The total 
solubility, ammonia, free acids, and alcohol were estimated 
in an aquous extract prepared by digesting finely crushed 
koji with cold water for 12 hours; the albuminoid nitrogen 
was determined in a precipitate thrown down in a decoction 
of fresh koji in in 40% alcohol to which a solution of cupric 
sulphate with some cupric acetate and afterwards sodium 
hydrate was added in such a quantity as to leave still a good 
trace of copper in solution. All other calculations were made 
from air-dry koji prepared by gradually transferring from a 
weighed portion small doses into a hot water stove to prevent 
any action of the ferments while drying. In the finely 
powdered air-dry material we determined the glucose and 
maltose in extracts prepared by repeatedly boiling weighed 
samples in alcohol of 80%. All other analytical methods 
were the same as mentioned in bulletin No. 2 with reference 
to the feeding stuff's. Difficulties only arose in the determi¬ 
nation of the dry matter in the powdered air-dry koji, which 
continually diminished in weight, even after 40 hours of 
drying. This was due, as we proved by special trials, to a 
continual disengagement of acid vapours consisting probably 
of butyric acid. Hence we took as the final results the 
