180 EXPERIMENT STATION EECORD. 



pp. 377-392, chart J).— Over 90 samples of bottled milk from New York City 

 were examined in order to study the factors which influence the distribution 

 of bacteria in the bottle. Samples were taken with the pipette, the milk 

 dipper, and by iiouring from the bottle. The conclusions reached were as 

 follows : 



"The upper 2 oz. of the cream of fresh bottled milk of fair quality contain 

 on the average 50 to 100 per cent more bacteria than an equal amount of the 

 lower cream. In older and more grossly contaminated milk the lower cream 

 may embody as many as or even more bacteria than the upper layers. By 

 removing these two top ounces from a milk bottle and using the remaining top 

 milk (8 oz. ) for infant feeding, as Hess has suggested, there generally results a 

 reduction of from 30 to 50 pei"- cent in the bacterial count. 



" The dominant controlling factor in the primary disposition of bacteria in 

 a milk bottle is the upward 'rafting' activity of the fat globules. A higher 

 percentage of bacteria are brought to the surface layers in a milk rich in cream 

 than one poor in that substance. At ice-box temperature the rate of increase 

 of bacteria in the cream and that in the skim milk are practically identical. 

 As the temperature is elevated the rate of multiplication in the skim milk out- 

 sti'ips that of the cream until at 30° C. it may be many times as rapid. In 

 certain samples of rather highly contaminated milk the abrupt change in the 

 temperature of the environment from 5 to .30° C. caused a striking bacteriolysis 

 in both the cream and the skim milk. This was probably an expression of 

 bacterial antagonism. 



"The sediment portion of the average bottle of fresh milk contains frequently 

 fewer bacteria than any other region of the fluid. A marked excess of bacteria 

 in the sediment indicates that the milk is old or that it has l)een kept in a 

 warm place." 



New and improved tests of dairy products, S. M. Babcock and E. H. Fae- 

 RiNGTON (Wisconsin »S7«. Bill. 1!)~), /)/>. ,3-1.3, figs. 5). — Details are given for 

 preparing and using a fat-saturated alcohol in the Babcock test. This is 

 placed on the top of the fat in the test bottle to aid in a more accurate read- 

 ing when testing cream. Readings with different test bottles before and after 

 using the alcohol showed that the accuracy of the test was increased from 

 0.5 to 2 per cent. 



A milk sediment test is described, which can be used for testing the different 

 lots of milk as they are received at the factory or receiving station. One pt. 

 of milk is poured into a cylinder and filtered through a thin disk of absorbent 

 cotton at the bottom. This disk contains the dirt and setliment from the milk 

 and can be dried and produced as evidence as to the cleanliness of the milk. 

 The cylinder is kept heated by a hot-water jacket in order to facilitate the 

 rapidity of the filtering. The success of the filter will dei)end largely on the 

 texture of the cotton disk, which should be made of absorbent cotton that 

 contains no starch or sizing. The records obtained by this test are an aid in 

 grading milk by its dirt content, and thus furnish a basis of paying for it 

 according to certain standards that may be agreed upon. 



An illustrated description is also given of the Wisconsin hydrostatic cream 

 balance previously noted (E. S. R., 22, p. 2S0). 



Reduction-fermentation test, O. Jensen (N. Y. Produce Rev. and Amcr. 

 Cream.. 29 {1910), Tslo. 23, pp. 858, 860). — A translation of an article previously 

 noted (E. S. R., 22, p. 478) on the method of using the reduction-fermentation 

 test for judging milk in paying according to quality. 



The alcohol milk test, J. H. Monrad (N. Y. Produce Rev. and Amer. Cream., 

 29 {1910), No. 25, pp. 938, 9JfO). — This is a description of the alcohol test for 

 determiuiug the quality of milk at the weigh can, as the reduction and fer- 



