374 
DR. P. E. SHAW ON THE NEWTONIAN CONSTANT OF 
Table I. (continued). 
Position. 
L. 
R. 
Mean. 
T.R. 
Corrected 
mean. 
Range. 
III. Falling Temperature. 
Mean 
range. 
°C. 
°C. 
B 
177 
18-5 
137-3 
210-2 
209-8 
173-75 
173-55 
95-70 
172-65 
57-85 
A 
194-0 
194-0 
269-3 
231-65 
231-65 
95-75 
230-50 
57-50 
57-65 
B 
160 
138-7 
209-6 
209-4 
174-15 
174-05 
95-80 
173-0 
57-65 
A 
152 
198-0 
198-2 
265 • 3 
231-65 
231-75 
95-80 
230-65 
In and after column 4, all readings are expressed in millimetres. 
Column 1 shows the position of masses, M, M. The next two columns have the 
temperatures of the masses, M, M, and of the vacuum tube, respectively. Columns 
4 and 5 give the extreme left and right scale readings, as the beam swings. Column 6 
shows the arithmetic mean of the two preceding column readings, while column 8 
shows this value corrected for change, if any, in the tube reading (column 7). The 
rest is obvious, heat being applied as already explained. There are three sections in 
the experiment, I., steady cold ; II., steady hot; and III., falling hot. By combining 
I. and II. we obtain as temperature effect 
a — +0'8 x 10~ 5 per 1 ° C. 
By combining I. and III., we find a like result. 
It will be observed that the vacuum is high ; the pressure I'O fx is calculated from 
the damping of the oscillations. 
Only this one example for cylinders will be quoted in full, for, though the readings 
are steady, the subsequent work with spheres is more sensitive and more reliable in 
general. 
In Table II. are entered the results of an experiment with spheres. Here we have 
seven sections, the last half, hot and cold, being taken ten days after the first. It 
will be seen that the interval of rest makes no appreciable difference in the result. 
Column 2 shows the time, column 3 the reading of the tube (reference) mirror. The 
fourth column has the extreme scale readings right and left as they occur, and the 
