388 Description of an Eudiometer. 
the residue of gas is expected to be very 
small, I employ a tube the sealed end of 
which is drawn out toa narrower diameter, 
so as to admit of more minute divisions (see 
fig. 6.) On the contrary, when only a small 
portion of gas is expected to be absorbed, 
the tube may be narrowest at the open end. 
To satisfy myself of the adequacy of this 
instrument to its purpose, I compared the 
analysis of artificial mixtures of oxygen and 
nitrogen gases, by its means, with that effected 
by nitrous gas used in Mr. Dalton’s mode ; 
by phosphorus ; and by detonation with hy- 
drogen. The results, in order to avoid all 
‘bias in favour, of any of the processes, were 
registered by Mr. H. Creighton, (to whom I 
am indebted for the annexed drawing) and 
when compared after the experiments were 
finished, they did not differ from each other 
more than 54, of the whole mixture. 
In graduating tubes for eudiometry or any 
other purpose, I have long been in the habit — 
of using a contrivance, which renders the 
operation greatly quicker, and insures perfect 
accuracy.’ It consists of a tube (Plate. VI. 
fig. 7.) open at both ends, and not more than 
.08 of an inch in diameter. ‘This is carefully 
divided into equal parts, which may be en- 
