1915] on Problems of Hydrogen and the Rare Gases 



553 



the mixture is known ; whereas in the former case of examination of 

 air unconden sable, this is uncertain, the only data being the pressure 

 remaining after exhaust and recondensation. But there is evidently 

 ample need for enquiry under different intermediate conditions in 

 order to throw more light on what appear to be new conditions of 

 relative volatility at low temperatures. 



DlFFCSIOX THROUGH PlATIXOI AND OTHER MeTALS AT HiGH 



Temperatures. 



The diffusion of hydrogen through a heated platinum tube can 

 ■easily be demonstrated. For this purpose a platinum tube A 

 (Plate 6) of J mm. bore and ^ mm. thickness of wall is fused up at 



Plate 6. 



■one end, while the open end is sealed into one limb of a glass 

 manometer B charged with sulphuric acid ; the platinum tube is 

 reduced to about 0'2 mm. thickness of wall for sealing into the 

 glass. It is further supported by an external glass tube (not shown 

 in the figure) fitting tight on to the horizontal extension of the 

 manometer, and drawn out to afford a good sliding fit on about 

 5 cm. of the platinum tube. The left-hand limb of the manometer 

 connects to a stopcock and charcoal bulb for exhausting between 

 ■successive measurements ; to facilitate this, the two limbs of the 

 manometer are connected by another stopcock C, which is closed 

 when a measure is to be made. For providing atmospheres of 

 different gases round the platinum tube, a quartz tube D of special 



