THE NEW ENGLAND STATES. 15 



detect any sophistication should it be attempted. The cooperation 

 with the Hydrographic Branch of the United States Geological 

 Survey was particularly fortunate because of their efficient field 

 force, consisting of men intimately acquainted with the local condi- 

 tions where the samples were to be obtained. Four-gallon samples 

 were collected in 2-gallon glass demijohns, which before being used 

 were subjected to special treatment to reduce to the minimum the 

 danger due to the possible solvent action of the water upon the glass ; 

 whenever a demijohn was found with a high coefficient of solubility 

 it was discarded. The sample was taken directly from the spring 

 whenever possible never from tanks or pipes if it could be avoided. 

 In the few cases where the water was covered by a cement or other 

 immovable cover so that the sample could not be obtained directly 

 from the spring, the fact has been stated in the description accom- 

 panying the analysis. As soon as the sample was obtained it was 

 sealed and sent to the Bureau of Chemistry for examination, where 

 the following determinations were at once made: Bicarbonic acid 

 (IIC0 3 ), nitrous acid (N0 2 ), nitric acid (NO 3 ), ammonium (NH 4 ), 

 hydrogen sulphid (H 2 S), carbon dioxid (CO 2 ), chlorin (Cl). The 

 sample was then divided into the several portions necessary for the 

 complete mineral analysis, and if for any reason the examination was 

 delayed these several portions were acidified with hydrochloric acid 

 to reduce to a minimum any action on the glass container by which 

 the composition of the sample, owing to the soluble constituents of 

 the glass, might be altered. 



The samples for bacteriological examination, collected by the 

 bacteriological chemist of the Bureau of Chemistry, were taken in 

 duplicate, packed in ice, and shipped to the bureau for examination. 

 A sanitary survey was also made by the bacteriological chemist at 

 the time of collecting these samples, and a full report was made on 

 the spot of the sanitary conditions of the spring and its surroundings 

 at that time. This fact should be borne in mind in considering the 

 statements descriptive of each spring. 



The methods of analysis employed throughout were in all essential 

 details those published by Hay wood, 1 with some slight modifications 

 when the peculiar character of a particular water or experience 

 suggested an advisable variation. 



The method for iodin and bromin devised by Haywood was used 

 throughout in the determination of these elements. A further study 

 was also made of the method, which emphasized the great delicacy 

 of the reaction, especially for iodin. It was found that 0.01 mg of 

 iodin in 100 cc of water could be readily detected without concen- 

 tration, and that by this colorimetric method, determinations of 

 one-tenth part per million of iodin were easily made. 



1 U. S. Dept. Agr., Bureau of Chemistry, Bui. 91. 



