R. R. Ramsey — Radioactivity of Spring Water. 311 



These observations were taken every minute and the mean 

 deflection from 15 minutes to 20 minutes from the time of put- 

 ting in the emanation was used. .By referring to an experi- 

 mental curve the maximum deflection per minute or the 

 deflection at the end of three hours was calculated. Correc- 

 tions were made for the natural leak and the amount of radium 

 C, from the former experiments. 



The method has an accuracy of 5 per cent or 10 per cent 

 as carried out at the spring using 15 or 20 minute readings. 

 When carried out in the laboratory using three hour readings 

 the accuracy is 2 per cent to 5 per cent. 



During March of this year the electroscopes were calibrated 

 by using Emanation standard E54, and were found at room 

 temperature, 20°, to agree with Duane and Laborde's formula 

 as given above to within 9 per cent. Emanation standard E54 is 

 certified by the Bureau of Standards to be accurate to within 

 3 per cent. 



If instead of using 6 - 31 X 10 6 as the value of the current in 

 electrostatic units which is given by one curie radium emana- 

 tion when in equilibrium with its products, one uses 6 - 02 X 10 6 

 as is done by A. Gockel,* the agreement is within 4 per cent. 

 Using the corrections for temperature and pressure as given 

 by the same author in the following formula : 



I = lobs. (0-0007(760-- p) + 0-002(« — 15°)) 



the agreement is within 3 per cent. 



Thus the Duane and Laborde formula under the above con- 

 ditions gives accurate results. 



The observations given below were for the most part made 

 at the springs within a half hour after taking the water from 

 the spring. The method of making the observation was as 

 follows : The shaking can is rinsed out and then filled by im- 

 mersing the can, if possible, and letting the water run in with 

 as little agitation as possible. When full the can is closed, set 

 level and then two stopcocks are opened, one at the top to 

 admit air and another on the side to allow the water to flow 

 out. When the water has quit flowing the stopcocks are 

 closed. By previous experiment the amount of water under 

 these conditions is known. The water and air is then shaken 

 vigorously for two minutes and then the air is pumped through 

 the electroscope until the emanation is equally distributed 

 throughout the air in the can and the electroscope. 



Readings are then taken on the leaf of the electroscope at 

 regular intervals for 15 or 20 minutes. Then the emanation is 

 pumped out. The observations taken at various times during 

 *Die Kadioaktivitat von Boden und Quellen, p. 102. 



