830 Messrs. A. Holmes and R. W. Lawson on 



drying-tube (c), and then through an arrangement (d) in 

 which are trapped any ions which may have been produced 

 in the process of shaking the solution. It consists simply of 

 an earthed brass cylinder of which the first part contains 

 several small wire-gauze filters, and the second part a central 

 insulated electrode charged by means of a dry pile (e). From 

 this condenser, the emanated air passes through the ionization 

 chamber (/) and thence to the water-pump past the mano- 

 meter (7i). The rubber tube connecting the manometer to 

 the water-pump is pinched by means of a screw slip, so that 

 a satisfactory working pressure can be obtained and main- 

 tained constant during the course of an estimation. The 

 ionization chamber consists of a central aluminium rod, con- 

 nected to the string of an Elster-Geitel electrometer and 

 insulated from the outer charged system. To prevent leakage 

 of electricity to the insulated rod, a guard-ring (i) is used. 

 By connecting the outer vessel of the ionization chamber to 

 a battery of cells (&), the other pole of which is earthed, the 

 necessary field for the production of the ionization current 

 is attained. The natural leak of the apparatus is first found 

 by passing an air current through a similar flask containing 

 an equal quantity of distilled water, and measuring the rate 

 of charging up of the electrometer. As in the measurement 

 with the solution, the distilled-water flask must be kept in 

 motion by the shaking machine and a constant pressure 

 maintained. This flask is then replaced by that containing 

 the solution to be tested, and the rate of charging of the 

 electrometer is again found. A known quantity of a standard 

 thorium solution is next added to the flask, and the operation 

 repeated. From the three readings so obtained it is a simple 

 matter to calculate the thorium content of the solution. 

 In the case of mineral solutions rich in thorium, only a 

 definite fraction of the whole solution was used, and the 

 total thorium calculated from the result obtained from this 

 fraction. For such measurements it is advisable to take all 

 readings over the same range of the electrometer scale ; for 

 instance, in the present estimations the current was measured 

 by noting the time required for the string of the electrometer 

 to pass over the first ten scale-divisions. The results obtained 

 for the various minerals examined are included in the table 

 of the following section of the paper. 



§ 4. Experimental Results. 



The minerals, analyses of which follow, are all from the 

 Langesund-f jord district south of Ohristiania, and are all of 



