ON THE MEASUREMENT OF THE LUNAR DISTURBANCE OF GRAVITY. Ill 



instrument. Now it appears that one complete revolution of a certain 

 tangent-screw by which the back-leg is raised should tilt the pendulum- 

 stand through almost exactly half a second of arc, and therefore this 

 should produce a relative displacement of the pendulum of the same 

 amount. We have no doubt but tliat a tenth of the turn of the tangent- 

 screw produces quite a large deflection of the image, and probably a 

 hundredth of a turn would produce a sensible deflection. Therefore, 

 from mere consideration of the effect of the back-leg wc do not doubt 

 but that a deflection of the pendulum through a ^liotli of a second of 

 arc is distinctly visible. This affords a kind of confirmation of the 

 somewhat unsatisfactory deductions which we draw from the operation 

 of the disturber. 



Posfscrijyf. — The account of our more recent experiments was written 

 during an absence from Cambridge from July 29 to August 9. In this 

 period the gradual southerly progression of the pendulum-bob, which was 

 observed up to July 28, seems to have continued ; for on August 9 the 

 pendulum was much too far S. to permit the image of the gas-flame to 

 come into the field of view of the telescope. On August 9 the image was 

 recentralised, and on the 9th and 10th the southerly change continued ; 

 on the 11th, however, a reversal northwards again occurred. Durino- 

 these days the unsteadiness of the image was much greater than we have 

 seen it at any time with the new instrument. There was some heavy 

 rain and a good deal of wind at that time. We intend to arrange a scale 

 for giving a numerical value to the degree of unsteadiness, but at present 

 it is merely a matter of judgment. 



It seems possible that earthquakes were the cause of unsteadiness on 

 August 9, 10, and 11, and we shall no doubt hear whether any earth- 

 quakes have taken place on those days. 



After August 11 we were both again absent from Cambridge. On 

 August 16 my brother returned, and found that the southerly progression 

 of the pendulum-bob had reasserted itself, so that the image was again 

 far out of the field of view. After recentralising he found the image to 

 be unusually steady. 



This appeared a good opportunity of trying the effect of purely local 

 tremors. 



One observer therefore went into the room and, standing near the 

 instrument, delivered some smart blows on the brickwork copino- round 

 the ditch, the stone pavement, the tub, and the large stone basement 

 underneath the water. Little or no effect was produced by this. Very 

 small movements of the body, such as leaning forward while sitting in a 

 chair, or a shift of part of the weight from heels to toes, produced a 

 sensible deflection, audit was not very easy for the experimenter to avoid 

 this kind of change whilst delivering the blows. To show the sensitiveness 

 of the instrument to steady pressure we may mention that a pressure 

 of three fingers on the brick coping of the ditch produces a marked 

 deflection. 



On August 17 1 returned to Cambridge, and noted, with my brother, 

 that the image had never been nearly so steady before. The abnormal 

 steadiness continued on the 18th. There was much rain during those days. 



On the afternoon of the 19th there was a high wind, and although 

 the abnormal steadiness had ceased, still the agitation of the image was 

 rather less than we usually observe it. 



The image being so steady on the 1 J'th, we thought that a good oppor- 



