140 Mr Parkin, On a reserve carbohydrate, 



extracted several times with cold water, till the mucilaginous 

 matter was dissolved out. The extract was filtered to remove 

 particles in suspension, such as starch grains ; then evaporated 

 down to a small bulk, refiltered and finally poured into a large 

 excess of alcohol, when a quantity of white flocculent matter 

 separated out. This was collected, dissolved in water and re- 

 precipitated by alcohol. Then treated similarly a third time, and 

 finally dried on a porous plate. About 1'5 grams of a translucent 

 almost colourless tough solid were thus obtained. This substance 

 swells and dissolves slowly in water to form an opalescent solution, 

 which does not give any marked colour with iodine. On account 

 of the opalescence its behaviour to polarised light could not be 

 studied with any degree of accuracy. Dilute solutions indicated a 

 dextrorotary power. 



A definite quantity, 0*959 gram, of this substance dried till 

 constant in weight, was dissolved in water, and the volume made 

 up to 100 c.c. Of this 25 c.c. were taken and boiled for one hour 

 with 2 per cent, sulphuric acid. The hydrolysis is not so quickly 

 effected as in the case of inulin, but one hour's boiling with this 

 strength of acid sufficed to complete it ; alcohol ceased then to 

 cause any precipitation, and the opalescence had quite disappeared, 

 leaving a perfectly clear solution. This hydrolysed portion first 

 showed levorotation, gradually changing to the opposite sign, and 

 after standing a few hours the rotation became constant. 



Angle of rotation in 200 mm. tube when constant = + o, 23. 



The cupric reducing power of the hydrolysed liquid was esti- 

 mated by ascertaining the weight of copper obtained from the 

 cuprous oxide formed on boiling with Fehling's reagent. 



Weight of copper calculated for the 100 c.c. = 1'7 grams. 



These two results do not agree at all closely on the assump- 

 tion that the sugar is glucose (dextrose), arising from the hydrolysis 

 of dextrin. The 1*7 grams of copper are equivalent to 0954 gram 

 of glucose, which should give a rotation at least four times as 

 great as that found. A second experiment with another portion 

 of the solution revealed the same discrepancy. 



The feeble dextrorotary power and the fact of the opticity 

 being at first left-handed, suggest at once mannose which exhibits 

 this form of multirotation. If it is this sugar, then, the character- 

 istic hydrazone should be deposited in the cold on the addition of 

 phenylhydrazin and acetic acid — a reaction which distinguishes 

 mannose from all the other hexoses. 



Some of the hydrolysed liquid, previously neutralised with 

 calcium carbonate, was concentrated on the water-bath, filtered to 

 free it from the calcium sulphate deposited, and then added to a 

 mixture of phenylhydrazine and acetic acid. After standing a 

 short time a bulky yellowish amorphous precipitate settled down. 



