Perkins — Molecular Weight of Radium Emanation. 4c6?> 



turns of German silver resistance wire, of less than 5 mm dia- 

 meter, were wound. After filling the spaces between the wires 

 by another coating of talc and water glass, this was followed 

 by a coat of rope asbestos 2 cra thick ; finally by an outside 

 coating of asbestos packing l - 5 cm in thickness. Ends for the 

 furnace were cut from thick asbestos board. 



Temperature was registered by a mercury thermometer. 

 The city current of 110 volts gave about 4*5 amperes through 

 the furnace and heated it to 250°C. in 20 min. After heating, 

 by aid of the city current, to within a few degrees of the tem- 

 perature required, the furnace was connected with the storage 

 battery, a sufficient number of cells being used to maintain the 

 temperature needed. An adjustable resistance of 14 ohms in 

 circuit was used to keep the temperature constant. After 

 equilibrium had been established, which usually took nearly 

 an hour, it was found that the temperature did not vary more 

 than - 5° per hour. 



Diffusion apparatus. — The diffusion apparatus must be 

 designed so as to minimize temperature changes and viscosit} 7 

 effects. The diffusion chamber must be air-tight except at 



the plug, which should be as symmetrically placed as possible 

 with regard to the diffusion chamber. It is necessary to insure 

 diffusion" through the plug in only one direction and to use 

 some suitable means for collecting and condensing the diffused- 

 vapor. 



Preliminary experiments were made with an apparatus con- 

 structed in glass including several stopcocks. This proved to 

 be impracticable owing to my inability to find a good lubricant 

 at these temperatures. 



The accompanying figure (tig. 1) is a section of the furnace 

 containing the diffusion apparatus as it was finally constructed 

 in iron. C is a piece of iron tubing about 30 cm long and of 2 ein 

 outside diameter on which collars with perfectly true edges 



