Appendix to Lecture of Jan. 22, 1915 815 



exhausted manometer* or McLeod gauge. A charcoal liquid air vacuum 

 arrangement serves for conveniently producing and renewing the 

 exhaust in the manometer ; but in the case of the McLeod gauge the 

 molecular pump is more suitable. The tube B is soldered gas-tight 

 into a coned and threaded boss C forming the centre of the closed 

 end of A. The hollow brass cylinder D completely encloses A, on to 

 the boss of which it fits tightly by its coned neck. A nut E secures 

 the arrangement against subsequent differences of pressure. The 

 rim of D is slightly coned inside to a similar hollow cylinder F, which 

 thus completes a gas-tight box enclosing the membrane. A second 

 lead tube G- similar to B opens into F and leads to a two-way stop- 

 cock, whereby the box is exhausted and filled to the required pressure 

 with the dried and purified gas to be examined. The tube G is bent 

 in a half circle and brought up parallel to B, so that the whole 

 arrangement can be enclosed in a bath at any required temperature. 

 For the use of higher pressures the two halves of the gas box seen in 

 (b) fit together with a spigot joint filled with a thin lead washer, and 

 are securely bound by peripheral screws on the external ring-plates 

 H and J. Five or six discs of coarse dried filter paper are placed 

 under the indiarubber membrane and form a readily permeable 

 support. The 2 mm. opening of B is covered by very fine mesh 

 copper gauze under the filter paper. The drum supporting the 

 membrane is only 2 or 3 mm. deep. This arrangement was tested to 

 forty atmospheres pressure. The pressure was regulated by a screw 

 valve attached to the bottles containing the compressed gases. The 

 gas passing through the membrane under high pressure appeared as a 

 slow stream of bubbles issuing from B, and could be collected and 

 measured in the ordinary way. At low temperatures and pressures the 

 gas transference was measured, in this case also, either by the increase 

 of pressure shown on an exhausted U-tube manometer of known 

 volume connected to B, or by a McLeod gauge when necessary. 



The rates of diffusion measured at a pressure of one atmosphere 

 with the apparatus of Fig. 1 gave the following values at 1.5° C. :^ 



Kelative Kates ^^^^ expressed in c.c. 



* See Fig. 9 : Problems of Hydrogen and the Rare Gases. Proc. Eov. 

 Inst. XXI., p. 558. 



