AND RATIO OF THEIR ATOMIC WEIGHTS. 



37 



oxyo-en in the globe iii the desiccator was very ueai'ly the same as that of the 

 hydrogen in the standard globe, the tube e was fused off. The pressure of the 

 oxygen was then reduced by drawing off mercury from the stopcock g, till (-(luilib- 

 liuni was produced at the differential manoraetei'. TIk^ water was well stirred, and 

 the e(|uilibrium was maintained for a sufficient time, sometimes for four hours, 

 commonly for one ht)Ui-. When the e(iuilibrium was regarded as satisfactory, the 

 stopcock of the globe was closed, the water was drawn off from the tank, the cou- 



FlG. 15.— Killing globe with oxygen ; use of desiccator; hydrogen used for comparison. 



nections of the globe to the other part of the apparatus were severed, the desiccator 

 was lifted out of the tank, dried, and taken to the balance. The cover «, Fig. 13, 

 was removed, the tube c was unscrewed, the ground joint of the globe was cleaned, 

 and the globe and its counterpoise were suspended from the pans of the balance 

 as before. 



25.— OXYGEN BY SECOND METHOD. THE BALANCE WHICH WAS EMI'LOVED. 



The balance used in the weighings of this series of determinations was made 

 by Rueprecht, of Vienna, and was lent to me by the Smithsonian Institution. It 

 will carry twelve hundred grammes iu each pan. The excursions of the beam are 

 read by the telescopic observation of the image of a scale of equal parts which is 



