330 PROCEEDINGS OF THE AMERICAN ACADEMY. 



I will now describe the method used in my work, and then take up 

 the experimental data obtained. As pairs of non-miscible liquids, I 

 have taken chloroform and water, benzol and water ; and as consolute 

 liquids, ethylalcohol, methylalcohol, and acetone. The next point was 

 how to determine the composition of the saturated solutions. The 

 methods of quantitative analysis are useless in this case ; but the 

 problem is solved without difficulty by quantitative synthesis. In- 

 stead of making a saturated solution and analyzing it, I measured the 

 quantities required to make a saturated solution at the required tem- 

 perature. Definite amounts of the consolute liquids were put in test 

 tubes by means of a carefully graduated pipette ; varying quantities 

 of one of the non-miscible liquids were run in from a burette, and the 

 second non-miscible liquid added from another burette to saturation. 

 The test tubes were corked, warmed just above the temperature at 

 which the final readings were made, so that there should be a single 

 homogeneous liquid layer, and placed in a constant temperature bath. 

 If the tube clouds, it is beyond the saturation point; if it remains 

 clear, it is not up to it, the required value lying between the two. 

 By making a series of experiments one can bring the limiting values 

 very close together, and thus determine the saturation point with great 

 accuracy. The constant temperature bath was at 20° C. No correc- 

 tion was made for the amounts of the three liquids evaporating off 

 into the vapor space in the upper part of the test tubes ; but by using 

 different sized test tubes this space did not vary much, being about 

 five cubic centimeters, so that the error due to this may be neglected. 



The chloroform used (Squibb's) was treated with sodium bisulphite 

 solution to free it from acetone, washed thoroughly with water, dried 

 over calcium chloride and fractionated, twelve hundred grams going 

 over within one quarter of a degree. Kahlbaum's crystallized benzol 

 was recrystallized twice and fractionated to constant boiling point. 

 The ethylalcohol was dried over lime and copper sulphate and frac- 

 tionated. The lot used distilled within half a degree. Part of the 

 acetone (from Eimer and Amend) was converted into the bisulphite 

 compound, back again, dried over potassium carbonate and calcium 

 chloride, and fractionated. Another portion was treated direct with 

 calcium chloride and fractionated. I could detect no difference be- 

 tween the two lots. I tried to purify a sample of acetone from Cutler 

 Brothers, purporting to be manufactured by Merck in Darmstadt; but 

 it was so bad that I used none of it in my experiments. The methyl- 

 alcohol (from Kahlliaum) was dried over anhydrous copper sulphate 

 and fractionated. 



