572 T ho day, — On the Capillary Eudiometric Apparattis of 
no longer tails off. The tube of acid is then removed, and the acid is blotted 
from the end of the capillary tube and from the surface of the mercury. 
After an analysis acid is drawn in as far as the pyrogallate film has 
reached, and the contaminated acid washed out a few times. Finally, with 
clean acid the whole tube is washed out as already described. 
A strip of paper is now inserted round the edge of the free mercury 
surface in the reservoir (h), and, by drawing this across, a clean surface 
is left, over which an empty tube is inverted. By rapidly moving the test- 
tube up and down over the end of the capillary tube, short threads of 
mercury are drawn in, separated by bubbles of air. When the chain reaches 
about two-thirds of the way along the graduated tube the test-tube is 
removed and the chain drawn rapidly back nearly to the bulb ; in this way 
the excess of moisture in the tube is passed by the mercury threads. 
These are now slowly moved to the right and they drive the water in front 
of them. When the end of the chain has moved over a portion of the 
graduated tube it is drawn rapidly back again ; most of the water is left 
where it was. This rapid backward and slow forward movement of the 
chain is repeated, the forward movement carrying the chain farther, after 
the first few times, so that gradually the whole of the chain is expelled. 
Finally the end of the capillary tube and the mercury surface are blotted. 
Should the mercury still not move freely in the tube, 1 acid may be left 
in it for some time, or better still a strong solution of potassium cyanide. 2 
The latter should be carefully washed out afterwards with distilled water. 
Finally the tube is washed as described. If this does not suffice the tube 
probably needs removing and thoroughly cleaning with nitric acid, &c., 
or the mercury itself requires purification. 
To make an analysis the air is drawn in to a mark (Fig. i, c) made half- 
way up the vertical part of the tube. The sample-tube is removed and the 
air drawn back till nearly the whole of it is in the bulb. It is now pushed 
back again, the right-hand end is adjusted accurately to the graduation 
marked 50 and the position of the other end read ; or the adjustment may 
be approximate and the position of both menisci read. The air is now 
drawn again into the bulb, again adjusted, and its volume read, and in this 
way several readings of the volume are made to test their agreement. 
The initial volume of the air having been obtained in this way, it is 
moved along till the right meniscus reaches point d (Fig. 1). Potash solu- 
tion is drawn in to point a , and very slowly expelled, till the air comes 
in contact with the film of potash left on the walls of the tube, the left 
meniscus being at a. If a high percentage of C 0 2 is expected, the potash 
is drawn in fairly quickly a second or third time and again very slowly 
1 To test this, watch for the vibration of the right-hand meniscus, or compare the readings 
obtained for the volume of air in the tube when drawn to the left and when pushed to the right. 
2 I am indebted to Dr. A. Lapworth, F.R.S., for suggesting the use of this solution. 
