84 MINERAL WATERS OF THE UNITED STATES. 



ROCKBRIDGE ALUM WATER (No. 1223 I. & W.). 



The composition of this water as determined by the Bureau shows 

 that it is markedly weaker in mineral content than even the least 

 highly mineralized of the springs as shown in the advertised analyses. 

 This is very likely due either to a gradual change in the composition 

 of the spring since the advertised anatyses were made or to fluctua- 

 tions in the composition from time to time. The following letter was 

 received from the spring owner: 



It may or it may not be a fair sample you obtained. Our mineral water comes 

 from rain percolating through a mass of mineral rock, the constituent parts coming to 

 the surface during dry weather, the rain bringing it into the springs, so that during 

 a dry time the water is materially weaker than during a wet spell or just after the rain. 



Except for special orders, we mix the various springs before bottling, and the water 

 you analyzed may have been from the weaker spring. The published analysis made 



by Professor was made after the following preparation. He remained at the 



springs two months, taking water from each spring two or three times a week during 

 the entire time, and on this he made his analysis. 



THOMPSON'S BROMIN ARSENIC WATER (No. 1224 I. & W.). 



The authors were unable to obtain the published analysis of this 

 water. It is something of a misnomer to designate this a "bromin" 

 water. 



WEBSTER SPRINGS SALT SULPHUR WATER (No. 1225 I. & W.). 



The agreement between the advertised and Bureau of Chemistry 

 anatyses of this spring is as satisfactory as could be expected on two 

 samples taken at different times. 



III. SARATOGA WATERS SAMPLED AT SOURCE. 



Besides examining the miscellaneous waters described in Part JI of 

 this report, all of which were purchased upon the open market, the 

 authors also made a study of the springs at Saratoga, N. Y., as they 

 come from the ground. An agent of the Department visited Saratoga 

 and collected the samples. In practically all cases the springs were 

 walled up so that it was impossible to take samples directly from the 

 source. Such being the case samples were taken at the nearest outlet 

 of the spring. 



The samples were shipped to the laboratory at Washington, D. C., 

 in the bottles ordinarily used at the various springs, and when received 

 all samples from the same spring were mixed in a clean, large, glass 

 demijohn, and subjected to immediate analysis. The authors would 

 have preferred to ship these samples in 5-gallon glass demijohns, which 

 are more resistant to the action of the water than the bottles obtained 

 from the spring owners, but feared to do so on account of the danger 

 of the large amount of carbon dioxid in the waters bursting the bot- 

 tles. Since the samples were not in these bottles for any great length 

 of time, however, it seems hardly possible that any material amount 

 of the glass was dissolved. 



