OOTOBEB 30, 1908] 



SCIENCE 



595 



ring a larger number than one, and point- 

 ing out that then there was little difference 

 between the ionic theory and Professor 

 Armstrong's. 



Sir William Eamsay thought discussion 

 was futile in view of the fact that the com- 

 pounds formed by the electrons and the 

 parts of the molecules, e. g., the ions and 

 their electric charges, had not been consid- 

 ered. He had thought that measurement 

 of surface tension in mixed liquids might 

 solve the problem, and had made a large 

 number of such measurements, but the fig- 

 ures had been found absolutely inexplicable 

 by Van der Waals and himself. The reason 

 for this was to be found in the difference 

 in composition between the surface layer 

 and the interior. 



Dr. Findlay pointed out that although 

 Professor Armstrong condemned existing 

 theories he had nothing to offer in their 

 stead, and that he had no quantitative evi- 

 dence to support the views put forward by 

 him. 



Professor Donnan in some trenchant re- 

 marks pointed out that Professor Arm- 

 strong's views were antedated by those of 

 "Werner, Bruehl and Kohlrausch. 



Dr. Wilsmore remarked that according to 

 Armstrong's theory the conductivity of a 

 solution should vary as the square of the 

 concentration, which like other deductions 

 was contrary to fact. 



Professor W. J. Pope stated that Pro- 

 fessor Armstrong was really only trying 

 to harmonize the ionic theory with the 

 views of chemists, by picturing the process 

 of solution by formulae like the structural 

 formula used in organic chemistry. 



Sir James Dewar read a paper on "the 

 production of helium by radium." 



Quartz vessels and glass joints were used 

 throughout ; the radium was that which had 

 been so carefully purified by Dr. Thorpe 

 (70 mgrs.). The Crookes radiometer was 



used to measure minute pressures. Solid 

 mercury warmed up to —23° just again 

 starts the Crookes' radiometer, correspond- 

 ing to a pressure of %o X 10"° mm. The 

 radiometer had to be washed out with oxy- 

 gen prepared from potassium perehlorate. 



When a McLeod gauge filled with air was 

 attached to a tube cooled by means of 

 liquid hydrogen to condense the air, the 

 pressure was reduced to 0.015 mm. repre- 

 senting the uncondensable gas at 20° abso- 

 lute. On again filling with old air (rich 

 in residues) the pressure was 0.00051 mm. 

 which became reduced to 0.00002 mm. on 

 washing out with oxygen prepared from 

 potassium permanganate. Hence it was 

 concluded that helium and neon adhere to 

 glass in the form of a film, and that one can 

 be easily deceived in measuring low pres- 

 sures. 



When 5 milligrams of radium were con- 

 nected to the radiometer it (the radiom- 

 eter) became active after a few hours, 

 although an attached tube containing char- 

 coal was cooled by liquid air. The amount 

 of helium produced by 70 mgrs. of radium 

 was measured in this way, the gases pro- 

 duced having to traverse a U tube filled 

 with charcoal and cooled by liquid air be- 

 fore reaching the McLeod gauge. After 

 one month's experiment a yield of 0.43 

 e.mm. of permanent gas per gram of ra- 

 dium per diem was obtained, but some of 

 the radium emanation had diffused into the 

 McLeod gauge and acted on some organic 

 matter or moisture, producing a higher 

 result. On heating up the charcoal, no 

 helium or other gas was given off. 



On repeating the experiment with all 

 care, and under the most favorable condi- 

 tions (any moisture or organic matter be- 

 ing now evidently eliminated), the radium 

 salt being kept one month in vacuo and 

 the glass containers constantly heated, 

 0.37 c.mm. of helium per gram of radium 

 per diem was obtained. This corresponded 



