EEV. T. E. EOBINSON ON SPECTEA OE ELECTEIC LIGHT. 
941 
the body to give a length of some inches, it was inserted in a piece of quill tube, whose 
lower extremity was fused on platinum wire reaching up to it ; and this was inserted in 
the discharge-tube. In tubes the distance of the electrodes was about 0‘75, in open air 
O'lO. The discharge-tube was cemented in a cap screwed to the air-pump, or rather to 
a transfer piece. This is a cylinder of brass with a concave screw above, a stopcock 
below screwing to the air-pump, and a lateral one connected with a desiccator. The 
desiccator is a bottle 8 inches deep and 2 inches diameter, fitted with a ground stopper, 
in which are two apertures. In one of these is cemented a tube which is connected with 
the transfer ; in the other, one descending to the bottom of the bottle, where it is drawn 
into a capillary point, its upper end being connected with either another desiccator or a 
gasometer. The desiccator is filled nearly with sulphuric acid. Supposing the transfer 
and the tube exhausted, shut off the air-pump cock and cautiously open that which 
connects the desiccator ; gas bubbles up slowly through the acid and fills the transfer 
and tube. Shut off the desiccator, connect the pump, and exhaust. By repeating this 
process ad libitum all traces of air or any gas previously used are removed, and nothing 
is present but the subject of experiment. It is almost needless to say that every part 
of this apparatus, including the air-pump, must be absolutely* air-tight. To ascertain 
whether this means of desiccation is sufficient, I used Kater’s oat-beard hygrometer. In 
the air of the room it read at 60° 3*358 R. It was then placed under a receiver con- 
taining a capsule exposing 20 square inches of sulphuric acid for five hours at a pressure 
of 0*3 inch, when it read 0*455 R. The receiver was then filled with air drawn through 
the desiccator as rapidly as could be done without drawing acid into the pump. In 10*" 
it read 0*285 R. Admitting air and introducing a slip of bibulous paper moistened with 
water, in lO*" it was 8*480 R; and when left in the air of the room some hours it was 
again 3*258. It is therefore evident that in this respect nothing more can be desired. 
Sulphuric acid also absorbs sulphurous acid and nitric oxide ; but when other acids 
might be present, a second desiccator was used filled with concentrated solution of 
potassa. 
The gasometer was made of a Woulfe’s bottle holding 75 cubic inches. In its central 
neck is cemented a siphon, one branch of which reaches to its bottom, the other to the 
bottom of a bottle of rather larger capacity; in the other two necks, stopcocks are 
cemented, one of which (A) is connected by elastic tube with the desiccator, the other 
(B) with any apparatus for generating gas. Suppose the Woulfe filled with water (or on 
a smaller scale with mercury). Let gas be supplied to B, the water will be displaced by 
it through the siphon into the bottle ; then closing B and opening A, the transfer and 
its tube are filled. If the water has been boiled for a few hours and cooled without 
exposure to the air, I find that gas continues in this gasometer sensibly pure for a much 
longer period than the duration of a day’s observations. For operating on vapours free 
from any mixture of permanent gas, I use a mercurial apparatus, consisting of a strong 
Woulfe’s bottle in whose three necks are ground (I) a glass stopcock, (2) a tube in which 
* That which I use has kept for a month a vacuum of 0*1 inch without variation of 0*01. 
6 N 2 
