i66 



NATURE 



[June i8, 1891 



the Common Balance." By J. H. Poynting, D.Sc, F.R.S., 

 Professor of Physics, Mason College, Birmingham. 



In a paper printed in the Proceedings of the Royal Society, 

 No. 190, 1878, an account was given of some experiments 

 undertaken in order to test the possibility of using the common 

 balance in place of the torsion balance in the Cavendish 

 experiment. The success obtained seemed to justify the 

 continuation of the work, and this paper contains an account of 

 an experiment carried out with a large bullion balance, in place 

 of the chemical balance used in the preliminary trials. The 

 work has been carried out at the Mason College, Birmingham. 



The Principle of the Experiment. — The immediate object of 

 the experiment may be regarded as the determination of the 

 attraction of one known mass on another. If two spheres, of 

 masses M and M', have their centres a distance d apart, the 

 attraction is, according to the law of gravitation, GMM'/^'-, 

 where G is the gravitation constant. Astronomy justifies the 

 law in certain cases as regards W/iP, but does not give the 

 value of G or M, except in the product GM. To find G we 

 must measure GMM'/</2 in some case in which both M and M' 

 are known. Having found G, we may determine the mean 

 density of the earth, for, assuming that it is a sphere of radius K, 

 the weight of any mass M' at its surface is 

 G X f7rR3AM7R2 

 = ^G7rRAM', 

 But if g is the acceleration of gravity the weight of M' may be 

 expressed as M'^. Equating these values, we get 



GttR" 



Method of Using the Common Balance. — With the length of 

 beam used (about 123 cm.) a differential method was applicable, 

 in which the attraction on the beam was eliminated. Two 

 spherical masses of lead and antimony, about 2i kilos, each, 

 were hung from the two arms of the balance, so that their centres 

 in the first position were about 30 cm. above the centre of a 

 large attracting mass, a sphere of lead and antimony about 153 

 kilos., placed on a turntable, so that it could be brought in turn 

 immediately under either of the suspended attracted masses. A 

 balancing mass of half the weight, and at double the distance 

 from the centre of the turntable, was found necessary, so that 

 the centre of gravity should be in the axis of rotation. Before 

 this was used, the ground level was seriously altered by the 

 rotation of the turntable. The attraction of the balancing mass 

 was calculated and allowed for. 



The alteration in the weights of the attracted masses, due to 

 the motion of the attracting masses from one side to the other, 

 was the quantity to be measured. When this was determined 

 in the lower position of the attracted masses they were raised to 

 about double the distance, and the attraction again determined. 

 The difference eliminated the pull on the beam, suspending 

 wires, &c. To lessen the effect of want of homogeneity or 

 sphericity in the masses, or of want of symmetry in the turn- 

 table, the masses were all inverted and changed over each to the 

 other side, and the weighings repeated. 



The position of the beam was determined by the reflection of 

 a scale in a mirror used with "double suspension." The mirror 

 was suspended by two silk threads, one attached to the end of 

 the ordinary pointer about 60 cm. below the central knife 

 edge, the other parallel to it, being attached to a fixed support. 

 The mirror turned through an angle about 150 times as great 

 as that through which the beam turned, and one scale division 

 corresponded to an angle of tilt in the beam of about 2/l5ths of 

 a second. 



The value of a scale division was determined by the use of 

 two equal riders which could be placed on or taken off wire 

 frames representing the scale pans of a small subsidiary beam, 

 2 "5 cm. long, fixed parallel to and at the centre of the large 

 beam. When one rider was placed on one supporting frame 

 the other was at the same instant lifted off the other frame. 



The balance was left free throughout a series of weighings, 

 and no moving parts of the apparatus were connected with 

 the case. 



The values obtained are as follows : — 



6-6984 



Mean density of the earth A = 5 "4934. 

 In the paper a description is given of a new form of 

 cathetometer used to measure the diameters of the masses. 

 NO. II 29, VOL. 44] 



The gravitation constant G 



••Quadrant Electrometers." By W. E. Ayrton, F.R.S,, J. 

 Perry, F.R.S., and W. E. Sumpner, D.Sc. 



In 1886 it was noticed, on continuously charging up the 

 needle of Sir William Thomson's bifilar suspension quadrant 

 electrometer No. 5, made by Messrs. White, of Glasgow, and 

 in use at the laboratories at the Central Institution, that the 

 deflection of the needle, when the same P.D. (potential differ- 

 ence) was maintained between the quadrants, instead of steadily 

 increasing, first increased, and then diminished ; so that, both 

 for a large charge on the needle as well as for a small, the 

 sensibility of the instrument was small. A similar effect had 

 been described by Dr. J, Hopkinson, in the Proceedings of the 

 Physical Society, vol. vii. Part i, for the previous year, and the 

 explanation he gives of this curious result is, that if the aluminium 

 needle be below the centre of the quadrants, the downward 

 attraction of the needle, which varies with the square of the 

 needle's charge, increases the pull on the bifilar suspen- 

 sion, and so for high charges more than compensates for the 

 increased deflecting couple due to electrical action. On raising, 

 however, the needle of our electrometer much above the centre 

 of the quadrants, the anomalous variation of sensibility of the 

 instrument with increase of charge in the needle did not dis- 

 appear ; and even when the needle was raised so that it was very 

 close to the top of the quadrants, and when, if Dr. Hopkinson's 

 explanation were correct, the sensibility (or deflection correspond- 

 ing with a given P.D. between the quadrants) ought to have 

 been very great for a large charge on the needle, it was, on the 

 contrary, found to be small. 



The needle was carefully weighed, with the platinum wire 

 attached and the weight dipping into the acid, and a calculation 

 was made as to the magnitude of the effect that should arise 

 from the change of the pull of the fibres due to any upward or 

 downward attraction of the needle by the quadrants. This 

 calculation showed that for a P.D. of 3000 volts between the 

 needle and the quadrants, the amount of such attraction was 

 quite unable to account for the observed diminution of sensi- 

 bility with large charges in the needle. Dr. Hopkinson says 

 in his paper, "Increased tension of the fibres from electrical 

 attraction does not therefore account for the whole of the facts, 

 although it does play the principal part." The experiments 

 that we made at the end of 1886 and beginning of 1887, con- 

 firmed by the calculation above referred to, proved that, at any 

 rate in our specimen of the quadrant electrometer, the principal 

 part of the anomalous action was not caused by an increased 

 tensionof the fibres, and that therefore some other cause must 

 be looked for to explain the observed results. 



We therefore decided to make a complete investigation of the 

 laws connecting the variation of the sensibility of the instrument 

 with the potential of the needle, the distance between the fibres, 

 the distance between the quadrants, and the position of the 

 needle. 



The results of the investigation, briefly summed up, are as 

 follows : — 



(i) The quadrant electrometer, as made by Messrs. White, 

 although it may be carefully adjusted for symmetry, does not 

 usually even approximately obey the recognized law for a quad- 

 rant electrometer when the potential of the needle is altered. 



(2) The peculiarities in the behaviour of the White electro- 

 meter are due mainly to the electrical action between the guard 

 tube and the needle, and to the slight tilting of the needle that 

 occurs at high potentials. 



(3) By special adjustments of the quadrants of the White 

 electrometer, the sensibility can be made to be either nearly in- 

 dependent of the potential of the needle, or to be directly pro- 

 portional to the potential, or to increase more rapidly than the 

 potential of the needle. 



(4) By altering the construction of the instrument, as de- 

 scribed, the conventional law for the quadrant electrometer is 

 obtained without any special adjustment of the quadrants beyond 

 that for symmetry, and the instrument is rendered many times as 

 sensitive as the specimen we possess of the White pattern. 



Linnean Society, June 4. — Prof. Stewart, President, in the 

 chair. — After nominating as Vice-Presidents Mr. A. W. Bennett, 

 Dr. Braithwaite, Mr. F. Crisp, and Dr. St. G. Mivart, the 

 President took occasion to refer to the loss which the Society had 

 sustained by the recent death of a Vice-President, Prof. P. 

 Martin Duncan, F. R. S. His genial presence at the meetings, no 

 less than his valued contributions to the publications of the 

 Society, would, he felt sure, be missed by everyone. — Sir Walter 

 Sendall, who was present as a visitor, exhibited a curious cocoon 



