BY H. S. HALCRO WARDLAW. 829 



telephone was used as the null instrument, and the readings of 

 resistance was made on a Wolff's pattern Wheatstone-bridge. 

 The conductivity-vessel was made with vertical electrodes to 

 hinder any settling of particles upon them. The cell-constant of 

 the vessel was determined by measurement of the resistance of 

 N/50 KCl solution. The determinations were all made in a 

 thermostat at 25°C. Readings of the bridge could be made 

 within a range of 1 in 200, and as the determination of the cell- 

 constant was subject to about the same variation, the values 

 obtained for the conductivities are subject to an error of 1 in 100. 

 Total solid s. — The percentage of total solid matter in the 

 samples of milk was determined by evaporating 1 cc. to dryness 

 in a watch-glass at 102^^0., in a glycerine-oven. The heating- 

 was continued for about IS hours. The solids were cooled over 

 CaClo in a desiccator before weighing. The amcjunt of milk 

 used was measured from a pipette which had been calibrated by 

 weighing, the calculations being made on the weight of milk 

 delivered. It has been shown by Poetschke (191 1), that the 

 weights of successive portions of a sample of milk, delivered from 

 the same pipette, agree together very closely. This result was 

 also obtained in the present case, when the extreme difference 

 between the weights of five portions of a sample of milk, de- 

 livered from the pipette used for the analyses, was 0"0004gm., 

 or 0'04%. The extreme variation observed in two series of 1 2 

 test-determinations each, of the weight of total solids obtained 

 from Ice. of milk, as described above, was 0*0016 gm. The 

 results obtained in the remaining determinations are assumed to 

 be comparable within this limit of variation (0"7%). The solids 

 of milk are very hygi-oscopic, when nearly dry, and it has been 

 shown by Leavitt (1910) and by other workers that, although 

 closely agreeing values may be obtained for the percentage of 

 moisture in substances like this when the estimations are carried 

 out under uniform conditions, it is a matter of some difficulty to 

 determine the absolute amount of water present. Further, de- 

 hydration of lactose proceeds slowly at temperatures of about 

 100° (Soldner, 189G). 



