in the Earth's Surface Materials. 763 



infer that the evolution of the emanation by the ebullition is 

 so rapid and complete that the whole of it, sensibly, is brought 

 into the space above the solution. It is probable that the 

 sensitiveness would be enhanced if the air withdrawn from 

 the electroscope was circulated again through the flask ; a 

 modification which does not appear to present any great 

 difficulties. Another important deduction from the experi- 

 ments cited above is that the presence of considerable quan- 

 tities of substances in solution does not appear to affect the 

 evolution of the emanation. The " rock-solution " contained 

 10 grams of a lava and the accompanying reagents ; the result 

 quoted is obtained as a difference on the thorium content of 

 the rock. These inferences have been substantiated by many 

 other experiments carried out for purposes of calibration. 



Under the conditions affording the values given in the first 

 table, we see that a gain of one scale-division per hour in the 

 rate of discharge of the electroscope corresponds to 2*8 X 10 -5 

 grams of thorium in the solution. It is quite possible under 

 ordinarily favourable conditions to be sure of a change of one 

 scale-division per hour by reading the normal rate of dis- 

 charge both before and after the test of a solution. If, now, 

 we deal with 50 grams of rock, the determinable quantity of 

 thorium is about 0*6 X 10~ 6 grams per gram of rock. In 

 order to make a comparison with the average amount of 

 uranium in rocks we may assume the case of a rock having 

 4 x 10~ 12 grams of radium per gram. Remembering that 

 3*7 x 10~ 7 grams of radium are in equilibrium with one gram 

 of uranium, we find that the amount of uranium present is 

 1*2 X 10 -5 grams. It follows that if there is as much thorium 

 as uranium in rocks, we may evaluate the mass of thorium to 

 a fractional part or deal with quantities of material much less 

 than 50 grams. 



Preliminary Results. 



The readings of the electroscope can be satisfactorily 

 standardized only by use of a thorium mineral of known 

 composition. The following few determinations, which are 

 based on the emanating power of Kahlbaum's thorium nitrate, 

 may therefore require some amendment later on. As com- 

 mercial salts have been found to possess an activity which 

 for the quantity of thorium present is less than that observed 

 in minerals, the correction on the present results would be 

 one of reduction. 



The solutions used had been in many cases originally pre- 

 pared for radium determinations. The quantity of dissolved 

 rock is, therefore, small. The readings were, however, per- 

 fectly definite and unmistakable. The major limit recorded in 



M 



