Royal Astronomical Society. 315 
ture was insufficient to point out the source of the anomaly: red- 
hot balls, lamps, or lumps of ice, placed near the torsion-box, 
failed to give any clue to the cause of the difficulty. 
All the observations, good, bad, or indifferent, made since the 
removal of the effect of radiation, have been taken into account in 
the general mean; and they have all been printed at full length, 
in such a manner that any one can go through the whole of the 
calculation in every experiment, from the announcement of one of 
the raw data of observation at the telescope, to the deduction of 
the mean density of the earth. Thus, as no experiment was ever 
more honestly or diligently performed, so also none has ever been 
more completely or satisfactorily described. 
Now as to the result. ‘The mean of all the experiments gives 
5°675 as the mean density of the earth, with a probable error of 
‘004. The balls used at the ends of the torsion pendulum have 
been lead of different sizes, brass, platina, zinc, glass, and ivory ; 
and in some of the experiments, a torsion rod of brass, without any 
balls, has been used. Various modes of suspension have been em- 
ployed, single copper wire, double iron, double brass, and double 
silk thread. The discordances which occur between the results of 
different balls, modes of suspension, or both, have every appearance 
of being the consequence of an insufficient knowledge of the torsion 
pendulum, and lend no countenance whatever to the suspicion that 
the attraction of matter upon matter varies in different substances, 
A moderate examination will show that there is no doubt that 
the discordances, being such as might have been looked for in 
any inquiry, must disappear in the mean of so large a number of 
observations. 
We may, then, confidently assert that this important element of 
the physical part of astronomy is settled, within very narrow limits 
of error. But suppose, if possible, that a less degree of trust were 
to be accorded to the mere result; nay, go further, and imagine 
the theory of gravitation itself, the best demonstrated of all general 
laws, to be an unfounded delusion. Perhaps we are then, on such 
a supposition, to give a still higher degree of praise to the manner 
in which this inquiry has been conducted. All the experiments are 
published, and all the experiments are facts. Ihave no doubt, 
and those who hear me have no doubt, that attraction is as real an 
existence in physics as it is an explanatory hypothesis in mathe« 
matics ; and experiments of the nature of that before us seem to me, 
as to you, to put this beyond doubt, both in the hands of Maskelyne, 
Cavendish, Reich, and Baily. But if we be wrong, how shall we 
ever be brought to know our error? How, except by experiments 
conducted with that firm honesty of purpose, and true absence of 
all bias, which has characterized those described by me today. I 
repeat that these experiments are facts, facts which are and will be 
true, facts which are the result of an inquiry in which the Newtonian 
doctrine was fairly thrown into the scale, and weighed by Nature’s 
own weights. ‘lhe confirmation of the general truth of Cavendish’s 
