MP:TH()DS of P]XAMIXATIOX 2(11 



is made, the milk should be well mixed. In milk that has 

 been thoroughly chilled, it is difficult to again mix the 

 cream equall\-, so one must be very exact in taking the 

 sample. It often happens, in the milk shops, that the 

 milk IS not well mixed when sold and an examination 

 of what is left may show so low a percentage of fat that 

 one immediately becomes suspicious that adulteration 

 has taken place, although it may not l)e so. 



b. Preserving the sample for analysis. If the milk sam- 

 ple is to be kept or shipped, it is necessary to add a pre- 

 servative. For this, potassium dichroinatc (I grni. to 1 

 liter) is used, which keeps the milk fluid foi- a long thne ; 

 this addition causes an increase in the specific gravitv 

 and the result of the chemical examination is also af- 

 fected. This must l>e taken into consideration at the 

 final test. It is generally better to add 20 drops of forma- 

 lin to every liter, yet in this ease it is possible that the 

 sj)ecific gravity of the wliey may be affected, on account 

 of a i)artial splitting of the casein. It is still better to boil 

 the filled bottle and then close it with a sterilized coik. 

 In taking the sample, it is also necessary to axoid ailding 

 water even in the slightest quantity (rinse the l)ottles 

 with milk), in order that no nitrate shall be added. 



c. The herd sample. In cases in which adulteration is 

 suspected, comjiarison may be made of the milk under 

 suspicion with that of the herd milked under super- 

 vision. The inii)ortance of the stable, or herd test has 

 been much overestimated. The great variations in the 

 composition of the milk of individual animals, leads one 

 to be careful in putting confidence in a comparison of the 

 fat content of milk on different days ; but comparisons 

 may fairly be made between the (luantities of solids not 

 fat and the specific gravity of the whey, which are fairly 

 regular. The stable or herd test is very uncertain, 

 therefore, for individual cows, or small herds, but it may 



