300 R. T. BAKER AND H. G. SMITH. 
Theory requires for this formula— 
b 
arbon ... 39°56 per cent. 
Hydrogen ... 7°69 a 
Oxygen vice BATH - 
From the results of the above tests, it is certain that the 
crystallised substance from this grass manna is mannite. 
Quantitative determinations of the Mannite.—A portion of the 
manna was dissolved in water, alcohol and acetate of lead added, 
a good precipitate forms, which easily separates, the filtrate being 
quite clear. From the filtrate the lead was removed by sulphur- 
etted hydrogen, the solution evaporated to dryness and boiled out 
with 90% alcohol, filtered, evaporated to dryness, and weighed. 
In this way 60°57 of mannite was obtained. As the greater 
portion of the sugars were by this method probably present, it 
was again dissolved in boiling alcohol, evaporated down, precipi- 
tated by ether, filtered, dried, and weighed. By this method the 
percentage was 58-98 of mannite. As the decomposition of the 
sugars by fermentation with yeast is slow, and perhaps not satis- 
factory, (see determination of the fermentation of the manna 
below), the method of determining the mannite gravimetrically 
by decomposing the sugars by yeast was not undertaken. AS 
the total amount of the sugars present is only just above 3% 
there could not be less than 57% of mannite present in the manna 
taken for analysis. As it is, perhaps, hardly possible to determine 
with accuracy the gravimetric value of mannite in mixed material 
like the present, we consider that we are justified in stating the 
approximate amount of mannite in this grass manna as 58 per cent. 
The gum was determined in the usual way. 
The substances insoluble in water were found to equal 27°58 
per cent. and consisted partly of debris, broken grass, and sub- 
stances of that character, partly of a small portion of carbo- 
hydrates allied to the gums, but largely consisting of a smooth 
light dirty-drab coloured mucilaginous substance that prevented 
the ready filtration of the material and which was practically an 
insoluble ferment. It has been described earlier in the paper- 
