903 



A study of tbe in contaminated wells of this district showed them to 

 contain a constant but comparatively large proportion of chlorine (10.7 

 to 11.3 parts per million). The proportion of nitrogen as nitrates in 

 these waters varies from 4.4: to 5.2 parts per million during the year ami 

 averages 4.7 parts. Taking the average drainage of the Harpenden 

 district as 8 inches, we have an annual loss of about 8 pounds of nitric 

 nitrogen per acre. Turning now to the contaminated wells we find the 

 chlorine ranges "from 11 (the standard of purity) to 173, the nitric 

 nitrogen from 5 to 44 parts per million of water." This variation is a 

 ready means of detecting contaminated wells. It is determined largely, 

 however, by the rainfall. "When once the soil has become satnrated 

 the eftect of any heavy rain is sure to appear in the well waters, but 

 usually about 1 month after the rain has occurred." The proportion of 

 chlorine to nitrogen varies greatly, but in individual pure wells, the 

 proportion is much more constant than is the ease with contaminated 

 wells. Tabulated data for the mineral constituents of Avater from 

 contaminated and iucontaminated wells are discussed as follows: 



There is a much larger (luantity of dissolved matter in the polluted than in the 

 pure waters. Tht; proj)ortiou of silica shows no rise with pollution; the proportion 

 of carbonates shows a very small rise. On the other hand the amount of lime, and 

 still more of magnesia, present in the waters receiving sewage contamination, is 

 much increased. * * * 



The great rise in the proportion of sulphates is, however, a fact which we have 

 not yet noticed; it is greatest in the nitric well, a circumstance which certainly sug- 

 gests the thought that both nitrates and sulphates are originally derived from the 

 albuminoids of food. It would appear that the quantity of alkalis in the con- 

 taminated waters was three or four times as great as iu the iucontaminated; the 

 alkali present was probably soda. The undetermined matter will consist in part of 

 alkalis; these were undoubtedly present, as the total acids were in all cases in excess 

 of the lime and magnesia determined. 



The quantity of suli)huric acid in the i)ure well water appeared so remarkably 

 small that a second determination was made later in the year, using two liters of 

 water for the experiment; the auiount of sulphuric anhydride obtained was, how- 

 ever, only 2. 75 per million. This amount is scarcely more than that found on an 

 average in the Rothamsted rain, namely, 2.52 per million. 



DIVISION OF STATISTICS. 



Eeport Xo. 95 (NEW SERIES), May, 1892 (pp. 151-190).— This 

 inchules articles on the condition of winter grain; progress of cotton 

 planting: changes in cultivated area; a wool retrospect; the French 

 tariff; reciprocity treaties with Salvador, Germany, Nicaragua, and 

 Uonduras; European cro^) report; and transportation rates. 



