O'Reilly — On the Barometer. 
33 
cause a disturbance of the state of equilibrium, and therefore to cause 
the column to fall away from its original vertical position ; and the 
extent of angular derangement of the column would necessarily be 
proportional to the extent of rise of the mercury. The mercury tend- 
ing to fall, the column would tend to regain the vertical, and reassume 
its previous state. But, as the column would thus tend to fall indiffe- 
rently to the one or to the other side, it naturally suggests itself that 
an appropriate arrangement of the counterpoise, would cause the column 
to remain on one or other side of the vertical in its limiting position, 
and thus oscillate entirely on that side of the vertical. Again, in 
order to measure the extent of oscillation of the column from this limit- 
ing position, there is merely requisite an index needle attached to the 
basin, and which may be made to combine with the counterpoise so as 
to determine the initial position of the column as regards the vertical. 
The arcs described by the extremity of this index needle on a graduated 
scale, appropriately situated, will then measure the extent of the oscil- 
lation, and therefore can directly indicate the corresponding barometric 
pressure. 
Such being the general principles of arrangement of the proposed 
system as represented by the model, there remains to be shown the 
particular advantage which I believe may be insured by it, and to ex- 
plain such a more appropriate arrangement of the different parts as 
would allow of its being constructed so as to fully insure that advan- 
tage. 
It is evident that the sensibility of the instrument will depend — 
1st. On the total weight bearing on the axis of suspension ; 
2nd. On the length of the glass tube ; 
3rd. On the angle of limiting position as regards the vertical. 
The length which may be given to the index will, of course, influ- 
ence the distinctness of the readings. 
The total weight bearing on the axis should evidently be reduced 
as much as possible, !N'ow, as the essential element of weight in the 
instrument is the mercury, it is evident that by limiting the diameter 
of the column, and the consequent volume of mercury necessary to fill 
it, this object may be best attained. On the other hand, as by the 
very nature of the proposed method of reading the meniscus ceases to 
be of importance, it is evident that the diameter of the tube may be 
advantageously reduced. As, however, the glass tube must have a 
certain thickness independent of the interior diameter, it is clear that 
the limiting bore will be that which can be obtained with the least 
thickness of glass, capillarity being avoided. A tube of from five to 
eight millimetres interior diameter would, perhaps, be a convenient size 
to construct with. As, moreover, the length of the tube must be taken 
into consideration in the working, since it must admit- of a certain 
amount of handling, it is evident that a smaller diameter than five 
millimetres can with difficulty be employed for such a length. 
The diameter being thus reduced to a minimum, the volume of 
mercury is proportionally reduced, and consequently that of the cylin- 
R. I. A. PROC. — VOL. I., SKIt. 11., SCIENCE. 
