ON THE MEASUREMENT OF THE LUNAR DISTURBANCE OP GRAVITY. 123 



January 23 the stone stood 8*62 mm. higher than on June 2t), 1881. 

 The observations are uncorrected for the eifect of temperature on the 

 metal rods, but the fact that the readings from the two rods of different 

 metals always agree very closely inter se, shows that such a correction 

 would amount to very little. 



The changes produced in the height of the stone are, of course, 

 entirely due to superficial causes ; but the amounts of the oscillations 

 are certainly surprising, and although the basements of astronomical 

 instruments may be very deep, they cannot entirely escape from similar 

 oscillations. 



In his address to the mathematical section at the meeting of the 

 British Association at Glasgow in 1876, Sir William Thomson tells 

 us' that Peters, Maxwell, Nyren, and Newcomb^ have examined the 

 observations at Pulkova, Greenwich, and Washington, in order to discover 

 whether there is not an inequality in the latitude of the observatories 

 having a period of aboat 306 days. Such an inequality must exist on 

 account of the motion in that period of the instantaneous axis of rotation 

 of the earth round the axis of maximum moment of inertia. The in- 

 equality was detected in the results, but the probable error was very 

 large, and the epochs deduced by the several investigators do not agree 

 inter se. It remains, therefore, quite uncertain whether the detection of 

 the inequality is a reality or not. But now we ask whether it is not an 

 essential first step in such an enquiry to make an elaborate investigation 

 by a very delicate instrument of the systematic changes of vertical at each 

 station of observation ? 



We will next attempt to analyse the merits and demerits of the various 

 methods which have been employed for detecting small changes in the 

 vertical. 



The most sensitive instrument is probably the horizontal pendulum of 

 Professor ZoUner, and its refinement might be almost indefinitely in- 

 creased by the addition of the bifilar suspension of a mirror as a means of 

 exliibiting the displacements of the pendulum-bob. If this were done it 

 might be possible to construct the instrument on a very small scale and 

 yet to retain a very high degree of sensitiveness. We are inclined to 

 think, however, that the variation of the torsional elasticity of the sus- 

 pending springs under varying temperature presents an objection to the 

 instrument which it would be very difficult to remove. The state of 

 stress under which the instrument is of necessity permanently retained 

 seems likely to be prejudicial. 



Next in order of sensitiveness is probably our own pendulum, em- 

 bodying the suggestion of Sir William Thomson. We are scarcely in a 

 position as yet to feel sure as to its merits, but it certainly seems to be 

 capable of all the requisite refinement. We shall give below the ideas 

 which our experience, up to the present time, suggest as to improvements 

 and future observations. 



Although we know none of the details of M. Bouquet de la Grye's 

 pendulum actuating a lever, it may be presumed to be susceptible of 

 considerable delicacy, and it would be likely to possess the enormous 



' B. A. Report for 1876, p. 10. For ' Nysen ' read ' Nyren.' 



« Peters' paper is in Bull. St. Pet. Acad., 1844, p. 305, and Ast. JVacJi. vol. 22, 

 1845, p. 71, 103, 119. Nyren 's paper is in Mem. St. Pet. Acad. vol. 19, 1873, No. 13. 

 With regard to ]\Iaxwell, see Thomson and Tait's JVat. Phil. 2nd edit, part 1, vol. 1. 

 An interesting letter from Newcomb is quoted in Sir W. Thomson's address. 



