828 THE VARIABILITY OF COWS' MILK, i., 



vicinity of IS^C, a sudden increase of density, which has been 

 termed the thickening' of milk, takes place. Fleischmann and 

 Wiegner (1913) have shown that this change is due to the solidifi- 

 cation of the milk-fat, wdiieh causes a discontinuous change of 

 density. In the present case, all the determinations of density 

 were made at temperatures within 17°C. and 22°C. Within this 

 range of temperature, the coefficient of expansion is practically 

 independent of the fat-content {cf. Midler's Tables, 1892). The 

 pycnometer was weighed to the nearest 25 mg., a weight which 

 corresponds to a difference of density of 1 in 10,000. This differ 

 ence is, therefore, the limit of accuracy of the determination. 

 The pycnometer was standardised with water at 25°C., and the 

 densities arc those of milk compared with water at this temper- 

 ature. 



Depression of freezing-poin t. — The freezing- 

 points of the samples of milk were determined by the method 

 described in a previous paper (Wardlaw, 1911). The greatest 

 variation observed between duplicate estimations v.as 0'002"C. 

 The freezing point of water could be determined within the same 

 limits. The depression of freezing point (A) is the diffei-ence 

 between these two values; its possible range of variation is, there- 

 fore, 0'001°. As this variation is on a depression of the freezing 

 point of water of about 0'55°C., the results obtained are com- 

 parable to 1 in 125. Milk contains substances in suspension. 

 Its value of A is, therefore, slightly lower than that correspond- 

 ing to the actual amount of substances in true solution (Tezner 

 and Roska, 1 908). In the present work, only comparable values 

 of the depression of the freezing point are required, and no 

 attempt was made to obtain absolute values, to obtain which, 

 special apparatus and precautions are required (Kaoult, 1898). 

 The difference from the absolute values of the present results is 

 very unlikely to be more than 0"01°C. 



Electrical conductivit y. — The electrical con- 

 ductivities of the samples of milk were determined by the usual 

 method of Kohlrausch and Nippoldt (1869). The secondary 

 circuit of an induction coil was used as the source of current, a 



