104 HATCH EXPERIMENT STATION. [Jan. 



of the feed law, of which 736 samples have been tested, either 

 individually or in composite. This makes a total of 2,045 

 substances analyzed during the year, as against 1,875 last 

 year and 1,147 in the year previous. There have also been 

 carried on for the Association of Official Agricultural Chemists 

 investigations relative to the best methods for the determina- 

 tion of starch, pentosans and galactan in agricultural products. 



Chaeacter of Chemical Work. 

 Water. — Sanitary examinations of water have been car- 

 ried out, as in previous years, according to the Wanklyn 

 process, to determine its general fitness for domestic pur- 

 poses and for the use of live stock. 



Persons whose water supply is other than that of a city or 

 town system should use every possible means to guard it 

 against pollution arising from sinks, vaults and stables, or 

 from the entrance of surface water and animal and vegetable 

 matter. The latter, while not in itself highly injurious to 

 health, is objectionable, as it favors the rapid propagation of 

 bacteria and other micro-organisms. The detection of specific 

 disease germs in water is, however, not a function of the 

 chemist, but of the bacteriologist. 



Frequent cases of poisoning result from conducting drink- 

 ing water through lead pipe, and such a practice cannot be 

 too severely condemned, for the poison, once assimilated, is 

 very difficult to remove from the system. At least five 

 samples examined during the past year have shown its pres- 

 ence. Soft waters as a rule have a much greater solvent 

 action upon lead than hard waters. Wells and springs ought 

 to be thoroughly cleaned at regular intervals. 



It is of great importance that the utmost care be exercised 

 in taking the sample for analysis, otherwise the chemical ex- 

 amination, conducted under the most careful and exacting 

 conditions, is of little or no value. The quantity necessary 

 is two to three quarts, collected in a thoroughly cleaned and 

 well-rinsed glass bottle, stoppered with a new cork, over 

 which is to be tied a clean piece of cotton cloth. An air 

 space of about one inch should be left between cork and 

 liquid, to allow for expansion. In case of pond water, the 

 sample should be taken from below the surface, being care- 



