Chemical Analyses of Agricultural Materials. iii 



of the previous contents, e.g. spirits or paraltin. be used. The bottle 

 should be mixed with the water which is to be analysed before being- 

 filled. Bottles should be nearly but not quite filled and should for 

 preference be closed with a glass stopper. Failing glass stoppered 

 bottles, clean new corks should be used. lu order to get a fair sample 

 from any running supply, such as the water puuiped from a borehole 

 or that drawn in pipes from a reservoir, the water should be allowed 

 to run for about an hour before collecting the sample. 



If the sample is to be drawn from a stream or pond, the stoppered 

 or corked bottle should be immersed some little distance from the 

 bank and the stopper or cork withdrawn well below the surface. The 

 ^topper or cork should be tied over with a piece of clean cloth. 



In the case of wells, boreholes, and, as far as possible, springs, 

 the situations, depths, and nature of strata should be stated. 



Liquids other tlian Water- 

 In sampling liquid substances, care should be taken to thoroughly 

 stir the contents of each package or drum before taking a sample. 



Ingesta and Similar Toxicological Materials— 



The advice of a veterinary surgeon should, if possible, always l)e 

 sought Ijefoie a sample of this nature is forwarded for analysis, and 

 precise information be given as to the nature of the poison suspected. 

 In the case of arsenical poisoning of stock, portions of those organs 

 showing marked inflammation should be included with some of the 

 ingesta. These samples should always be most carefully packed in 

 a screw-topped fruit jar or bottle. 



Milk and Cream— 



Mixed. — In taking a sample of mixed milk from a number 

 ■of cows, the milk must be poured from one vessel to another several 

 times, and the sample must be taken immediately before the milk is 

 allowed to settle. If the sample is made up from mixed milk from 

 several vessels, it should contain the quantity from each vessel which 

 will ensure that the completed sample is an average of the whole bulk. 

 The stirring of the milk is in each case not sufficient. 



One Cow. — The cow must be stripped thoroughly, the milk 

 strained and well mixed by pouring from one vessel to another several 

 times, and the sample must be taken immediately after, before the 

 milk is allowed to settle. 



Cream samples should be prepared by stirring thoroughly and by 

 pouring the cream from one vessel to another several times, and the 

 sample must be taken immediately after. 



(Note. If it is desired to take composite samples of milk or 



-cream application should be made for full information and instruc- 

 tions, which will then be given.) 



