88 ANALYSIS OF MILK. 



to a porcelain dish and mixed with 15 to 20 grammes of 

 purified sand (obtained by iguitinj; " silver sand," boiling this 

 with 25 per cent, hydrochloric acid, and thoroughly washing and 

 drying). The use of such a medium as sand is essential in order 

 to secure intimate and complete contact between the reacting 

 substances at the drying point — which is the point of reaction. 

 The mixture is made alkaline with barium hydroxide, and 

 evaporated to dryness. Two c.c. of a 1 per cent, alcoholic solu- 

 tion of curcumin are added, and the mixture again evaporated 

 to dryness, the mass being stirred from time to time to ensure 

 thorough incorporation. To the mixture is now added 1 c.c. of 

 a solution containing 25 c.c. hydrochloric acid and 10 grammes 

 of oxalic acid in 100 c.c. of water, and the mass is again dried. 

 The same operations are carried out with 10 c.c. of a standard 

 solution of boric acid [1 c.c. being equal to O'l milligramme of 

 boron trioxide (B^Oj)]. 



The colour having been obtained in both cases, the sand 

 is extracted with ordinary alcohol. 



The coloured solution obtained from the milk is diluted with 

 alcohol or methylated spirit until the colour is of the same degree 

 of intensity as that formed from the standard ; and the amount 

 of boric acid is arrived at by an obvious calculation. 



The colours are compared in two tubes of the same internal 

 sectional diameter (about 1 centimetre) placed vertically against 

 a white porcelain plate. 



On comparing the two solutions it will occasionally be found 

 that a certain amount of orange tint is exhibited by one or the 

 other, due to the presence of curcumin in slight excess. When 

 this is observed the tints must be made the same by the cautious 

 addition of solution of curcumin to the solution which does not 

 show the orange tint. 



The result of a number of experiments made with known 

 amounts of boric acid and borates show that the process is 

 reliable and accurate. 



Estimation of Citric Acid in Milk. — The proteins are precipi- 

 tated by acid mercuric nitrate (p. 90) and a measured volume 

 of the clear filtrate exactly neutralised with dilute caustic soda 

 solution, using phenol-phthalein as indicator. A white precipi- 

 tate of mercury and calcium phosphate and citrate is thrown 

 down, collected on a filter, and washed with water ; it is removed 

 from the filter, suspended in water, and a little dilute liydi-o- 

 chloric acid added ; sulphuretted hydrogen is passed through to 

 precipitate the mercury as mercuric sulphide. After filtration, i he 

 the solution is boiled to remove sulphuretted hydrogen, and, after 

 the addition of a little calcium chloride, cooled. It is then care- 

 fully neutralised, phenolphthalein being used as indicator ; the 



