COMPOSITE SAMPLES 313 



tests 3.7 per cent the quantity of fat contained would be equal to 

 85 (pounds of milk) times 0.037 (test) equals 3.145 pounds 

 of fat. 



Composite Samples. — A composite sample is one made up 

 of a number of smaller samples. If a small quantity of the 

 milk yielded by a cow morning and night for a week be added to 

 a glass jar and adequately preserved, a single testing of the 

 composite sample at the end of the week will indicate the aver- 

 age quality of milk produced by that cow throughout the week, 

 provided the quantity of sample taken is in proportion to the 

 amount of milk given. This system is employed in whole milk 

 creameries where the milk delivered by each farmer is thoroughly 

 stirred and a small portion added to the test sample, which is held 

 for a period of from one to two weeks, and then tested. This 

 method is convenient and economical of time, but must be oper- 

 ated carefully or expensive errors may be made. 



A glass bottle, with large ground-glass stopper (Fig. 107) is 

 by far the most desirable sort of container for the sample. The 

 two points to be guarded in the holding of the sample are: First, 

 the prevention of evaporation and, second, the prevention of the 

 loss of fat due to its sticking in the crevices of a rough cover. 

 A sample kept even half a day in an open dish like a tea cup is 

 utterly worthless as a means of determining the fat content of 

 the milk produced or delivered, since so large a quantity of 

 water would have evaporated that the remaining portion would 

 be too rich in fat. For the same reason, but to a less degree, a 

 bottle with a large cork stopper is not to be recommended, some 

 water would escape through and some fat would cling to the 

 cork. Mason fruit jars have often been used. These prevent 

 evaporation, but cause more or less loss of fat about the top. For 

 ordinary farm use they probably will serve as well as anything 

 to be found. In the earlier days of creamery work, and to 

 some extent up to the present, pint milk bottles with metal caps 

 which clamp more or less tightly were employed. It has been 

 shown by trial that the amount of water which will evaporate 

 from such a lid is very appreciable, indeed, and that many cream- 

 eries are paying for more fat than they actually receive on this 

 account. This reduces the percentage over-run on the part of the 



