8 On the Mean Density of the Earth. 
experiment. Hence it appears that our result cannot be made 
to agree with that of Mr. Cavendish, unless our } 16 seconds be 
diminished to about 10°5 or 10:4, on the supposition of an error 
of more than a whole second in excess, in the number 11°6 se- 
conds; which cannot be admitted without doing great violence 
to the observations. 
Having thus failed in our endeavour to discover any error, ¢ or 
even suspicion of error, in the conduct or result of the Schihallien 
experiment; let us now turn our attention to the other experi- 
ment, as performed by Mr. Cavendish. And here I must at once 
disclaim all expectation of meeting any failing with regard to the 
operator himself, whom I well knew'to be a most excellent phi- 
losopher and mathematician, as well as a patient, accurate, and 
acute experimenter. The failure then, if any, must be expected 
from the nature of the machine, and of the calculations—From 
the perusal of Mr. Cavendisn’s account of the machine he employ- 
ed (in the Phil. Trans. of 1778, or vol. xviii, of my edition), and 
the nature of the arithmetical calculations, they at once appear 
. to be formidable and discouraging in the highest degree. The ~ 
machine is small, comparately with those in the former or moun- 
tain experiment. It is not easily to be understood, without ac- 
tually seeing it, though assisted with the view of the drawing of 
the whole, on account of the intricacy and perplexity of the con- 
struction. In the first place, at each end of a light wooden rod 
of near two yards in length, is attached a me AT leaden ball of 
two inches diameter; the middle of the rod being fixed to, and 
suspended by, a long and very slender copper wire ; by any smal! 
movement of these balls and the connecting rod, in a horizontal 
direction, by the torsion or twisting of the wire, a very minute 
and slow vibratory motion iscommenced. ‘To produce this small 
motion in the two little balls, and their connecting rod, two other 
large balls, of ten inches diameter, are connected together by 
certain machinery, at like distance -as the former, and capa- 
ble of being moved to different distances, or positions, on the 
horizontal level with the small balls. By so setting the large 
balls near the small ones, these are attracted by the former, pro- 
ducing a very small motion in them, and in consequence a very 
slow vibration. So minute are these motions, that the extent 
of the vibrations is but a small fraction of an inch, and the dura- 
tion of each vibration is not performed but in the time of several 
minutes, from 3 or 4 to near 15 minutes. So minute are these 
motions, that telescopes and other means are necessary to 
view and to estimate their quantity and durations. 'To produce 
these minute motions, very complex machinery is necessarily 
employed, while the delicate movements are watched for many 
hours together, during many days, and recorded with regard to 
the 
