1910. | with the State of Aggregation of Matter. 121 
and allowing the aldehyde to distil off as soon as itis formed. The latter, 
after separating from the aqueous layer, was dried over fused sodium acetate, 
and then fractionated with a Young rod and disc still-head and the fraction 
boiling below 120° was separately collected. This was converted into 
the bisulphite compound, which was washed with ether, and air-dried, 
decomposed with a small excess of bicarbonate and the aldehyde then 
distilled oft. To the aldehyde was then added, little by little, and with 
constant cooling, half the volume of ordinary concentrated ammonia, and the 
mixture, after standing a day, completely solidified. The valeraldehyde 
ammonia compound was then washed with ether, and air-dried. The com- 
pound from 300 grammes aldehyde was then treated in three portions with 
300 c.c. water, and to the suspension was added 150 c.c. anhydrous hydro- 
cyanic acid, diluted to 270 cc. with water. The cyanammonia compound 
was then hydrolysed, and the remainder of the preparation completed by 
the method of E. Fischer.* From 300 grammes of aldehyde were obtained 
200 grammes of twice recrystallised pure inactive leucine. 
The determinations of solubility were carried out by mixing 1 gramme 
(or more if 1 gramme were insufficient for complete saturation) of the 
crystalline substance with 30 c.c. of the salt solution in short wide-mouthed 
test-tubes, which were then tightly corked, and clamped on to a rotating 
axis in a thermostat, and thus, by means of a motor, kept in a state of 
constant motion, generally for a period of 36 hours. At the end of this 
time the liquid was filtered from the undissolved solid by means of a water 
pump through a perforated platinum plate covered with asbestos, into a 
flask with a narrow neck graduated in tenths of a cubic centimetre between 
20 and 30 c.¢., which was carefully calibrated before use, and which was 
provided with a side-tube near the mouth of the flask, turned upwards 
so as to connect with the water-pump for suction. The whole apparatus 
was immersed in the thermostat during filtration, and the amount of liquid 
filtered was directly read. 
The nitrogen was determined in this liquid by means of Kjeldahl’s 
method, except in the case of the nitrates and sulphocyanides. In both these 
cases the leucine and phenylalanine were estimated by Sorensen’s formal- 
dehyde titration method, and in determinations of the solubility of other 
substances in sulphocyanide the undissolved crystallised substance left on 
the filter-plate was dissolved in some organic solvent, and the nitrogen 
determined in this solution after filtering through blotting-paper. In the 
latter case there is a slight error, owing to the fact that it is impossible to 
wash the crystals. The Kjeldahl method of estimation was, in the case of 
* ‘Ber.,’ 1900, vol. 33, p. 2370. 
