210 THE CHEMICAL CONTROL OF THE DAIRY. 



and the composite sample is taken by pouring out from the 

 spout. 



Composite Samples. — It is sometimes convenient to test 

 the milk from a particular source once a week, or at otter inter- 

 vals ; in this case the specific gravities should be noted daily, 

 and a measured portion, such as 11 c.c. or 1 oz., placed each day 

 in a bottle containing a little solid potassium bichromate ; the 

 fat estimation is made on the mixed sample. 



Sample Cans. — The most convenient receptacles for the 

 samples from bulk are round cans about 2 inches in diameter 

 and 4J inches deep, or wide-mouthed glass bottles which can be 

 closed by a disc or stopper. In large dairies, the milk is received 

 from difierent farms, and it is convenient to stamp the name 

 of the farm legibly on its own sample can; sample cans should 

 be provided and marked for all the samples it is desired to take 

 regularly, and the use of unmarked cans, cans marked with 

 paper labels, and cans marked with another designation should 

 be avoided, if possible. 



The lids of these cans should fit well, and they should not 

 be filled more than three-quarters full, to obviate as far as possible 

 the risk of spilling in transit. The samples, after being taken, 

 should be placed in a box or tray made to contain twenty-four 

 cans, or any other convenient number, in which they can be 

 transported to the laboratory. The trays should have a strong 

 handle in the middle, and it is desirable that they be also fur- 

 nished with a lid which can be locked or sealed, to prevent any 

 tampering with the samples in transit. The duty of conveying 

 the samples should be entrusted to one man, who should be 

 made responsible for any spilling of the contents. A tray holding 

 twenty-four cans is not too heavy to be carried steadily, and 

 there is no reason why any of the milk should be spilt. If the 

 sampling is performed at a distance from the laboratory, the 

 use of bottles is more reliable ; a case to contain these should 

 be provided. It is, of course, necessary for the chemist to see 

 that the sample cans are in good repair ; an)' which leak, or 

 have ill-fitting lids, should be replaced at once. The cans or 

 bottles before being handed over to the samplers should be 

 quite clean and dry ; the cans should be washed with hot water 

 and dried, and when not in use, kept with the lids open. If 

 bottles are used a large stock should be kept, and they should 

 be well washed with warm water and allowed to drain for a 

 couple of days ; this will be found to dry them sufficiently. 



Testing. — The tests applied to milk should be of two kinds, 

 simple tests done in the dairy, and a more extensive examina- 

 tion in the laboratory. The only test sufficiently convenient 

 for use in the dairy is the estimation of the specific gravity. 



