Bonnier and Mangin for the Analysis of Air. 567 
Analyses are carried out rapidly enough for ordinary changes of 
barometric pressure to be negligible, and the apparatus is so constructed 
that all the readings for one analysis are taken under appreciably the same 
pressure of mercury. This is ensured by the horizontal placing of the 
measuring-tube and by the relatively large surface of the mercury in the 
reservoir H. 
It is, of course, necessary for accurate results that the mercury should 
move perfectly freely in the tube, and the first difficulty lay in attaining 
this condition. Even after the mercury had been carefully cleaned, by 
dropping through dilute nitric acid and finally by distillation in vacuo , 
a scum appeared when the tube was washed with water after hydrochloric 
acid, as recommended by Aubert, and this scum it was almost impossible 
to remove completely. I infer it to be calomel produced by the action of 
hydrochloric acid on traces of oxide in the mercury, 1 and precipitated when 
the film of hydrochloric acid is diluted with water. When dilute sulphuric 
acid was used instead of hydrochloric acid this difficulty disappeared, and 
the washing of the tube was greatly facilitated. 2 
When next the tube is thoroughly washed out with distilled water and 
* the excess of water is carefully driven out by a series of short mercury 
threads, the residual film breaks up into drops, and a mercury meniscus no 
longer moves smoothly, but with a series of jerks. This is avoided if the 
water used for the final washing contains some sulphuric acid ; I have 
therefore used latterly a 5 % solution of sulphuric acid as the sole washing 
liquid. The film then left on the walls of the capillary remains unbroken, 
and a mercury meniscus moves perfectly freely in the tube, vibrating about 
its position of rest before becoming stationary. The film also serves the 
purpose of ensuring a constant vapour tension. 
1 Berthelot : Comptes Rend., xci, p. 871 ; Barfoed : Journ. f. prakt. Chem. (2), xxxviii, 1888, 
p. 459. 
2 It is, of course, necessary to protect the open surface of mercury when the apparatus is not in 
use, and especially to keep the small amount of dust and oxide which collects on the surface from 
getting into the capillary. 
