1912 -13.] Determining Vapour Densities at High Temperatures. 5 
To avoid this source of error it is sufficient to mark the position of 
manometer, slit, and screen on the table. A displacement of any one 
of these by one or two centimetres has no appreciable effect. It is 
convenient to have the slit and lens securely fixed, as an alteration in 
the comparatively small distance between them produces the greatest 
effect. 
(2) If the screen is rotated about a vertical axis out of its position 
perpendicular to the ray of light, the relative distance of the images 
is increased in inverse ratio to the cosine of the angular displace- 
ment. This source of error is, however, very small. It is sufficient 
to place the screen by eye as nearly as possible perpendicular to the 
ray of light. 
(3) A rotation of the manometer about an axis perpendicular to the 
plane of the mirrors produces a displacement which is greater, the greater 
the initial distance between the images. 
When a point of light is substituted for the slit, the two point images 
should not he more than 1 or 2 mm. apart. It is quite easy to adjust 
the mirrors prior to sealing the membrane into the outer quartz cover G, 
so that this may be the case. A rotation of the manometer during a 
determination by more than an exceedingly small amount can be avoided 
by securing the quartz tube K (fig. 3) to a wooden support which is kept 
in position on the table by a heavy weight. 
When the above-mentioned sources of error are avoided, the zero point — 
i.e. the distance between the two images — is perfectly constant. As quartz 
is the only material used in the construction of the manometer, change 
of temperature does not cause a relative displacement of the mirrors. A 
good manometer should give a deflection of J mm. on the screen, with 
a pressure difference of 1 mm. of mercury, and should stand a pressure 
difference of 1£ atmospheres without breaking. With this degree 
of sensibility the pressure readings are reliable to less than 0’5 mm. of 
mercury. 
The apparatus for adjusting the pressure in the outer jacket G (fig. 3) 
is shown in fig. 4. 
When the three-way tap c is in the position shown in the drawing, the 
vacuum pump is in communication with the vessel a. By turning the tap 
c through 180° a is brought into communication with the atmosphere. 
The space above the mercury in b is nearly completely evacuated, but a 
little air is left in it to act as a cushion and prevent the top of the vessel 
from being broken by the impact of the mercury, should air be let too 
suddenly into a. 
