MECHANICAL STRETCHING OF LIQUIDS. 
361 
The branch LM, which is more than the length of the mercury barometer, hangs 
vertically. For the sake of greater clearness of representation the diagram shows the 
bulbs HK, CD, and AB with their axes in the same plane, but in practice it is more 
convenient when making the connection at E to turn the apparatus that is to be filled 
through a right angle about EF as axis, since in this position liquid contained in 
either bulb can be more easily boiled by means of a small Bunsen flame. 
The liquid to be experimented on is placed in a large beaker N over a layer of 
quite clean mercury 2 or 3 centims. deep, and into this beaker the open end M of the 
tube ZM dips, not quite reaching the mercury. After the liquid in N has been 
boiled to expel dissolved air, it is allowed to enter the bulbs HK, CD, and AB, and 
is then boiled in each simultaneously by suitable flames.* Thus, on the removal of 
the flame from beneath AB, air-free liquid from CD enters and fills it ; and when 
that in CD ceases to boil, air-free liquid from GHK enters and fills CD, ancl so on. 
The process of alternately boiling out and filling is continued, with pauses, for 
several hours, till the residual bubbles disappear very quickly on cooling, and the 
tendency to boil by bumping threatens to endanger the apparatus. 
Then while the liquid in HK is kept gently boiling, the tube LM is heated nearly 
to redness, just below L, in order to drive off attached air which is carried off in the 
stream of passing vapour. 
Then when this tube has again cooled, and the vessel HK is about half-full, the 
vessel N is raised so as to submerge the open end M below the mercury, which, when 
all is cool, rises to nearly barometric height. Then after any liquid that has condensed 
above the mercury in ML has been, by judicious heating, driven over into the bulb 
GHK, the tube LM is heated just below L to the softening point, and allowed to 
close under atmospheric pressure. This process of sealing invariably liberates gas 
either from the surface of glass or by decomposition of the vapour of the liquid. It 
appears, however, to be less in quantity when the tube near the sealing point has 
been previously subjected to strong heating in the manner described. 
The portion of the apparatus thus cut off from the barometer tube is now turned 
about a horizontal axis along GE, through 180°, into the position shown in fig. 3, 
Fig. 3. 
and by careful heating the liquid filling the tube EF is boiled over into the bulb HK 
till when all has cooled to the temperature of the room, the surface stands at some 
* It is important that the stem AG of the tonometer shall be provided with an enlargement at the 
end in order that the instrument may not burst when the mercury in the ellipsoidal bulb is raised to the 
temperature of the surrounding hot liquid. 
MDCCCXCII,—A. 3 A 
