The result of the analysis was obtained as follows, most operations being carried out in duplicate :—- 
per cent. 
per cent. 
Extract by petroleum spirit (boiling under 45 degrees C.) 
1-02 
1-10 
Extract by ether (water free) chlorophyll 
0-40 
0-39 
Extract by absolute alcohol, principally glucose, with a trace 
of tannin ; no alkaloids 
0-73 
Lost. 
Extract by cold water ... 
27-04 
As the previous treatment with ether and absolute alcohol made some of the albuminoids insoluble, 
watery extract of the flour was prepared. 
per cent. 
per cent. 
Extract by cold water of air-dried flour ... 
28-48 
27-96 
This watery extract contained : — 
Ash ... 
1-65 
1-60 
Total nitrogen 
1-054 
1-055 
Soluble albuminoids calculated from nitrogen 
6-59 
6-59 
Soluble albuminoids coagulated when boiling the watery extract, 
vegetable casein 
1-18 
Precipitated by absolute alcohol : — 
1st. Mucilaginous substances 
3-22 
314 
2nd. Dextrin 
4-98 
6-70 
When treating the watery extract with 4 vols. of absolute alcohol, there is a danger that part of 
the saponin is also precipitated with the dextrin, if the filtering is not done very quickly. As the second 
sample filtered very slowly, the dextrin contained a large amount of saponin, by giving the characteristic 
test, when treated with sulphuric acid. 
per cent. 
Not precipated by absolute alcohol, glucose ... ... ... ... ... 064 
For comparison with the usual method of extraction, extracts with benzol and 80 per cent, boiling 
alcohol were also made :— 
per cent. 
Extract by benzol (C 6 H 6 ) boiling at 81° C. ... ... ... ... ... 1-43 
Extract by 80 per cent, alcohol (Sp. Gr. -8483 at 15°) ... ... ... ... 19-88 
Of this extract was 
Soluble in abs. alcohol ... ... ... ... ... ... ... ... 2-50 
Insoluble in absolute alcohol, but soluble in water, and precipitated by sub¬ 
acetate of lead ... ... ... ... ... ... ... ... ... 17-16 
The saponin was determined in accordance with the method recommended by Christophson and 
Otten (Dragendorff’s Plant Analysis , p. 68), by extracting with boiling 80 per cent, alcohol, filtering when 
hot, boiling off the alcohol of this extract, and precipitating the saponin with concentrated baryta water. 
For the quantitative determination, the residue of the flour, after having been treated and extracted with 
petroleum ether, ether and absolute alcohol were used, and gave in an average of several determinations -- 
14-58 per cent, saponin. Larger quantities of the glucoside were prepared from fresh samples of the flour. 
The purest saponin was obtained on cooling of the alcoholic solution, in the form of a white curdy 
precipitate. I also separated the saponin by evaporating the alcoholic solution, precipitating with baryta 
water, and decomposing the baryta saponin with carbonic acid, and separating the saponin by shaking the 
watery solution with chloroform. 
The yellowish amorphous powder obtained had a peculiar sweetish taste, was easily soluble in water, 
chloroform, and in hot dilute alcohol. From the solution, it was precipitated with subacetate of lead, 
baryta water, and also, to a slight extent, by acetate of lead, but gave no reaction with Mayer’s solution. 
By moistening a trace of the dry saponin with strong sulphuric acid, or, better, with fuming sulphuric 
acid, a beautiful red colour is slowly formed, turning purple and becoming darker on standing. This 
colour keeps for days. A small quantity of the saponin dissolved in water forms a strong froth on shaking 
the solution. 
Both these reactions may be shown with slices of the fresh or dried seeds, as when moistening the 
surface with sulphuric acid, very shortly bright red spots and streaks are formed ; again, when putting a 
few slices of the seeds in water and shaking the mixture, a froth is formed as if soap were present. 
