792 
DR. T. R. ROBINSON ON THE DETERMINATION OE 
D 
V' 
P 
P-eV' 2 . 
20 
• 9-49 
0 - 464 
0-067 
29 
1436 
0-993 
0-047 
52 
21-50 
2-112 
0-079 
It will be observed that for the smaller cups e is relatively larger than in the pro- 
portion of their areas. This is, in great measure, owing to the arms, 0'5 inch diameter 
and 24 inches lonm bearing so much greater proportion to them. From the preceding it 
follows that = eV' 3 X 4 • 7 1 . 
As to what I call gyroscopic friction, for want of a better name, the pressure which 
produces it =(A — C) (a> / w /// ) ; A and C being the moments of inertia round the 
principal axes, and oj / m /// the angular motions round them. C was determined, but 
it would be rather awkward to get A. As, however, the angular velocities are as 
Y and v, the effect of this friction will be merely to lessen (3. 
(23.) 4. In the rotation of an anemometer at the end of a revolving arm, the 
centrifugal force produces an outward pressure which must be resisted by some 
stop, and produces there a friction f" . This pressure is given by the formula (C) 
F"=y^|jXY 2 (not V' 3 ), where M' is the weight of the parts which revolve round 
the axis of the anemometer, G the distance of the centre of gravity of M' from the 
vertical axis, and R the length of the horizontal arm. This pressure was in some of 
these experiments considerable ; in No. 133 it was 52 lb. ; on another occasion it broke 
a steel arm of No. Ill, fths of an inch thick. As the resulting friction varies nearly as 
V' 2 , it might, like f", have been included in a ; but as I hoped that the results of these 
experiments might be available for real wind (where f"' has no place), I thought best 
to measure it and add it to f I determined it thus : A strong upright was fixed in the 
gallery, and secured by a strut to the platform; to this was fixed a pulley 1*73 inch 
diameter, as nearly as the eye could judge in a line with the seam of the horizontal 
arm, which was my nearest guide to the direction of the axis of the shaft. Over this 
passed a fine iron wire attached to the centre of the shaft, to which weights were 
suspended which pulled it in the direction of its length. The friction thus produced 
was measured by weights placed in the cups, and the normal f subtracted from this. 
The tensions were corrected for the friction of the pulley, which when this line crossed 
it at right angles was - 2 ^th of the load. The only other mode of measuring this friction 
was the setting the arm vertical with the 12-inch disc below, placing weights on this 
and proceeding as in the case of f This, however, would have required the total 
dismounting of the apparatus and constructing a, proper stand, for which time could 
not be spared. 
With P'"=10lb., F=180 grains ; with P ,v = 21 lb., F = 280, each a mean of three 
F . 
trials. These give — , in the first instance=7 , 032, in the second=8 - 36I. It seems 
