THE CONSTANTS OF THE CUP ANEMOMETER. 
789 
increase of f— 9'7 grains for the added pressure = 0"294 grains for every ounce, or 4"7l 
for every pound. Hence for the different anemometers — 
No. I. /=113T9. 
No. II. „ 101-27. 
No. III. „ 201-32. 
No. IV. „ 184-98. 
No. Y. „ 118-53. 
(21.) 2. Brake friction . — This is produced by means shown in Plate 70, fig. 4. A is 
the box mentioned (2), to the bottom of which is screwed the strong brass frame b. 
In this turns the lever D, with its arbor E. The longer arm of this lever is connected 
with the piece F, slotted to give room for the shaft and one of its three friction 
wheels, as shown in the figure, and attached below to the wire W, which passes down 
through the vertical axis and its perforated step, and carries a scale-dish below. The 
shorter arm of D is connected by the link G to the right angled lever HH', whose 
arbor is at L. If now a weight be placed in the dish it draws down F, and thereby 
I), and G raises H, makes 1 1/ press the piece M, which turns with the arbor N, and 
presses the rubber R which it carries against the friction-disc C, six inches in 
diameter. A duplicate of this mechanism is placed opposite (but not shown, to avoid 
confusion), and being also actuated by the wire W, presses a similar rubber against the 
corresponding part of C. K is one of the counterpoises to balance the scale-dish, 
which, however, was much too heavy, as I had no idea what weight might be 
required, and it produced unnecessary friction. The rubbers are of stout cloth, 1"25 by 
3T 6 '5 
0'8 inch, and show no signs of wear. The leverage of the brake is — X ; _ = 8’03. By 
this elegant arrangement (for which I am indebted to Mr. Grubb, F.R.S.) the friction, 
while it can be varied at pleasure, causes neither vertical or lateral pressure on the 
supports of the shaft. When, therefore, the brake is made to act, it merely adds to f 
another friction, which can be similarly measured. The measures, however, of this 
friction were much more discordant than those of f : two differing in one instance 
234 grains. This irregularity is probably caused by the tremors of the ground 
(already referred to) acting on the brake-weight ; when thrown upwards it would 
relax its pressure momentarily, and permit the measuring disc to revolve with a less 
pull than what really represented the friction. 
When I measured this by Atwood’s process at Armagh, the results were very 
regular, but far larger than those obtained at Rath mines, except on one occasion when 
Mr. H. Grubb took a set very early in the morning before there was work going on. 
This agreed well with them. As, however, all the anemometer experiments were 
affected by this vibration, I deemed it best to use the Rath m ines frictions. 
But I found an unexpected fact : that the coefficient of this friction is not constant, 
but decreases with the pressure ; this is probably owing to the elastic nature of the 
