642 MILK 



be made with special care. In the first place a representative sample 

 must be obtained. Similar amounts should be removed from differ- 

 ent parts of the can, so that the top, the center, the bottom, and 

 the intermediate parts between the center and the periphery are 

 represented. Butter samplers are very suitable for this purpose. 

 Equal parts of each portion should be mixed and the ice-cream then 

 allowed to melt. The viscosity of the cream retains air-bubbles 

 tenaciously and, therefore, samples must be weighed and "not 

 measured. Food inspectors should always be careful to take 

 representative samples. Samples of cartons of ice-cream or from 

 small portions of a can may be misleading. 



SelHng ice-cream in stores at retail has sometimes been com- 

 pared to skimming milk. Siace under ordinary store conditions 

 the ice-cream is liable to become soft, there is a movement of the 

 fat to the top, and early customers obtain a cream richer in fat, 

 but poorer in heavy sohds, than do late customers. 



The bacterial content of ice-cream has been frequently found 

 to be very high. Stiles and Pennington found the following 

 numbers of bacteria in ice-cream in Washington : In 1906 and 1907 

 263 samples gave an average of 26,612,371 bacteria per cubic 

 centimeter. The maximum count was 365,000,000 and the mini- 

 mum count 137,500. Ayers and Johnson give the following table: 



. Slimmer. . ■ Winter. . 



Bacteria per cubic centimeter. No. of samples. Per cent. No. of samples. Per cent. 



Oto 50,000 .. 5 5 49 



50,000 to 100,000 . . 8 8 79 



100,001 to 600,000 9 9 57 23 25 27 



500,001 to 1,000,000 9 9.67 2 2 20 



1,000,001 to 5,000,000 16 17.03 21 23 08 



5,000,001 to 10,000,000 12 12 77 5 5 49 



10,000,001 to 20,000,000 13 13 83 11 12 09 



20,000,001 to 30,000,000 10 10 64 6 649 



30,000,001 to 60,000,000 8 8 51 3 3 30 



60,000,001 to 100,000,000 9 9 57 4 440 



Over 100,000,000 8 8.61 4 4.40 



Hammer gives the following results of investigations of ice- 

 cream : 



• Bacteria per cubic centimeter. . 



„,., ,, oourco. Year. No. of samples. Average. Highest. Lowest. 



Philadelphia 1905-6 49 17,833,031 79,800,000 70,000 



Bos'o.n 1906-7 35 23,000,000 150,000,000 1,000,000 



Washmgton 1906-7 263 26,612,371 366,000,000 137,600 



Chicago / ISO^ 89 16,662,134 125,000,000 20,000 



,,., °, 1 1910 306 15,401,000 100,000,000 20,000 



Milwaukee 1911 26 8,000,000,000 200,000 



pes Moines 1911 10 19,920,000 39,000,000 4,200,000 



Iowa State College 1911 12 19,775,000 72,000,000 600,000 



Ayers and Johnson have also studied the groups of bacteria in 

 ice-cream, with the following results: 



Summer. Winter. 



Per cent. Per cent. 



Arid-coagulating 49 82 30.84 



Acid-forming 20 72 38,03 



la^t • ■. 13.98 4.81 



Alkali-forming 1 86 6.42 



Peptonizing I3 52 20.90 



