546 PROFESSOR DANIELL ON THE WATER-BAROMETER ERECTED 
While the water in the boiler, which now constitutes the cistern of the baro¬ 
meter, was still warm, a quantity of the purest castor oil (Oleum Ricini), was 
poured into it till the surface was covered to the depth of half an inch; this 
was done for the purpose of cutting' off the communication of the atmosphere 
with the water, and with the view of preventing the absorption of the air. 
Some of the same oil was poured upon the surface of some distilled water in a 
wide-mouthed glass vessel, and being lightly covered with paper was set by in 
a closet, that any change might be detected to which it might be liable under 
such circumstances. 
The adjustment of a scale was the next object of importance. For this pur¬ 
pose a hollow brass rod (1,2) was prepared of ftlis of an inch diameter, and 
adjusted by means of a screw at the upper end to a flat ruler of brass (2, 3) 
divided into inches, and carrying a vernier (4) by which the hundredth part of 
an inch is easily read off, and which is moveable from the outside of the case 
of the instrument by means of a rack and screw (5). The same rack and screw 
also moves a brass screen (6, 7), which rises and falls with the vernier and pro¬ 
tects the tube from the heating influence of the breath or hand ; a small ther¬ 
mometer is inserted into this screen. The rod was measured from a scale for¬ 
merly belonging to the late Mr. Cavendish, and now the property of Mr. 
Newman, by marking it with a beam-compass at intervals of two feet, and 
afterwards repeating the process at intervals of sixteen inches. The two mea¬ 
sures corresponded to the one twentieth of an inch ; the difference being found 
to depend upon the multiplication of a small error in laying down the sixteen 
inches, and corrected accordingly. 
The rod was next placed in the case of the barometer by the side of the 
tube, being made to pass through the wooden stays of the tube, in which it 
can freely move. At its lower end an ivory point of known length was fixed 
by which it was very carefully brought into exact contact with the surface of 
the oil in the cistern ; the flat scale was then carefully adjusted to its upper 
end, and it was fixed at the lower end by screws to the top of the copper 
cistern. The column of water was thus found to stand exactly thirty-three 
feet four inches, or four hundred inches above the level of the fluid in the cis¬ 
tern. This, then, is the neutral point of the instrument, above or below which 
a correction of ±’02 inch must be made for every ascent or descent of five 
