= aghiy etait . re RT PSE rey Se BS EN Sy oe a ee Pee RR PRE ae he 
w eS erat ct 2 ire! ida 
445 
Two difficulties were met in the use of both the methods which I have 
described. At the high pressure of about 200 atmospheres which is 
here reached, the glass tubes are likely to explode upon the slightest 
provocation. Fortunately, although explosions in the external tube were 
frequent, the tube B within the bath rarely gave way. This difference 
between tubes of the same form and drawn from the same piece of glass 
is doubtless to be attributed to the annealing and toughening of the glass 
at the temperature of the bath. The danger of explosion in the outer 
_ tube was reduced as far as possible by choosing flawless tubes and drawing 
them with care. The capillaries never burst and probably could have 
withstood a much greater pressure than that to which they were subjected. 
On account of their flexibility and strength they form an excellent means 
of communicating high pressures. 
The other difficulty was that ordinarily enough silver oxide could not 
be packed into the tube B to give the pressure of equilibrium before 
being entirely decomposed. This difficulty was overcome by filling only 
one-half of the tube with silver oxide, the rest being filled with powdered 
potassium permanganate. The latter decomposes at once at the tem- 
perature of the bath and produces a much larger quantity of oxygen 
than the same volume of silver oxide. The two substances were separated 
by a small piece of asbestos. 
Experiments in which the manometer was used showed that equil- 
ibrium was reached in from fifteen to forty minutes, according to 
the kind of oxide used. In one case an addition was made to the appar- 
atus in an extra bulb at C filled with potassium permanganate. After 
the manometer had come to rest as usual, the bulb at C was gently heated 
causing a further increase of pressure, but this excess of pressure rapidly 
disappeared through the reoxidation of the silver in B and within ten 
minutes the pressure was the same as before, within a small fraction of 
an atmosphere. This shows that we are dealing here with a case of true 
equilibrium. 
The first experiments with both methods gave disappointing results. 
The pressures calculated with the aid of Boyle’s law, ranged between 
175 and 200 atmospheres. Those obtained by the indirect method were 
on the average several per cent higher than those obtained with the 
manometer. However, this difference disappears when allowance is 
made for the difference in compressibility’ of air and oxygen contained 
respectively in the tubes EG and HI. 
A more careful investigation of the temperature in the neck of the flask 
pointed to the cause of the observed variations in pressure. Although 
a pretaainary test bad shown the temperature to be the s same in the neck 
ee oi oa ae a oA weu. | 
ad we i 
