230 Mr GRAHAM on the Law of 



it should be ; but it is difficult to find elements for a proper 

 correction, as not only the quantity of vapour in the atmosphere 

 must be taken into account, but also the hygrometric state of the 

 plug itself. The increased return-air, however, evidently lowers 

 the diffusion-volume of the hydrogen gas. 



With the view of increasing the capacity of the instrument, 

 and the number of its divisions, and of obviating the interference 

 of vapour, the mode of performing the experiment was varied. 

 On a tube, four-tenths of an inch in diameter, a bulb of two 

 inches in diameter was blown, as in figures 

 A and B. The tube above and below the 

 bulb, in the case of A, was graduated into 

 two-hundredths of a cubic inch. The upper 

 end of the tube was closed by stucco, as in 

 the case of the simple diffusion-tube. The 

 general mode of proceeding will be best con- 

 ceived from the recital of the details of a 

 particular experiment. 



The diffusion-instrument employed in the following experi- 

 ment contained 855 measures, and was of the form A. The 

 stucco plug was unusually large, being 0.6 inch in length, which 

 occasioned the diffusion to be slow. At the commencement of 

 the experiment the thermometer stood at 68, and the barome- 

 ter 29.73 inches. The bulb being sunk in water with the air- 

 syphon in it, the whole air was withdrawn, with the exception of 

 12 measures, and the instrument filled up with newly made hy- 

 drogen gas. So that at the outset we had in the instrument, 



Air with its vapour, , ' ." : ..V*. * 12 

 Hydrogen, . . ' '. ' ; . ' . . 823.83 

 Vapour (accompanying the hydrogen at 68), 19-17 



855.00 



