56 R. T. BAKER AND H. G. SMITH. 
a little chlorine. Only a very small amount of calcium 
was detected and this was evidently due to the accidental 
presence of a few of the raphides which had passed through 
the cloth. It is thus apparent that the mucilaginous 
portion of this tuber consisted largely of the organic salts 
of potassium and magnesium, 
On the addition of an equal volume of 90% alcohol to the 
mucilaginous solution and shaking, a glairy mass separated 
in strings, which quickly floated to the top of the liquid. 
The filtrate was quite clear and bright, and on addition of 
two volumes of alcohol to this, and standing over night 
only a very small amount had separated; this had the 
character of a vegetable substance allied to arabin, but 
was too small in amount to determine with certainty. 
Moisture-10 grams of the ‘tuber’ cut through the 
centre, were heated at 100-105 °C. until constant; the solids 
weighed 0°4824 gram, so that the water present was 
95°176%. A duplicate determination gave 0°4825 gram 
solids. The dry substance was of a light brown colour and 
had an odour strongly resembling that of chicory, for which 
substance it would form a good substitute. It was 
difficult to prevent any portion becoming brownish when 
exposed to the air, or to heat, and the aqueous extract 
soon became slightly coloured, although it was colourless 
at first. The dried residue when ignited and fully car- 
bonated gave an inorganic residue equal to 1°276% on the 
tuber. 
Ash—As the ash of the above contained alumina (a very 
unusual thing with plant substances belonging to the 
Phanerogams), a portion of the ‘tuber’ was taken from the 
centre so that no possible contamination could take place. 
The total ash from this was 24°11 on the dried substance 
a little less than that of the whole tuber and alumina was 
present. This material was taken for a quantitive deter- 
