126 EXPERDJlvXT STATIOxN. [Jan. 



]yater Analysis. 



There is no one feature' of the home of greater importance 

 than a pure and unfailini;' water supply. In order that families 

 living where a i)ublic su])i)ly is not available may have the op- 

 l)ortunity of determining the i)urity of the water from the well 

 or spring they are obliged to nse, the experiment station will 

 make a chemical examination of the water at a nominal charge 

 ($3 a sample). Water for examination must be shipped in con- 

 tainers, wdiieh will be sent to the applicant on request, and the 

 blank forms sent with the coiitainer must be filled out and re- 

 turned with the samide. A bacteriological examination is not 

 made, and the chemical analysis is limited to those determina- 

 tions which will indicate sewage contamination or the presence 

 of lead or ol)jectionable metals. 



One hundred and ten samples w^ere examined during the past 

 year; of these there were 77 samples from wells, 28 from 

 springs, 4 from ponds and 1 sample of ice. A large number of 

 well w\aters examined w^ere condemned. Some showed only 

 slight contamination while others were pronounced dangerous 

 for use. 



Wells located close to dwellings and stables are particularly 

 liable to contamination ; hence great care should be exercised 

 in the location of new wells. Where a well is known to he free 

 from contamination it should be bricked or cemented up well 

 above the surface of the ground in order to prevent the entrance 

 of surface water during wet weather, and the top of the well 

 should be made vermin proof. 



Milk, Cream and Feeds for Free Examination. 



In addition to the other w^ork of this department, free analy- 

 ses have been made of 372 samples of milk and 1-iG samples of 

 feedstufFs. A large number of butter-fat tests on cream by the 

 Babcock method have also been made. 



As a general rule, it is preferred to have ap])lication made 

 before a sample is submitted for analysis. Directions for 

 proper sam]ding and for shi])ping can then be forwarded. An 

 analysis of an im])roperly drawn sample may lead to erroneous 

 conclusions. In the case of feeding stuffs an analysis is often 



