508 



Major J. Herschel. 



nexion with the theory of the torsion gravimeter, together with remarks 

 upon other forms of gravimeter. This now appears as Part I of the 

 present paper. 



So far as I was aware, the theory of the torsion gravimeter had not 

 been published. In considering it, I was led to think that in some 

 respects it involved peculiarities which might be taken advantage of 

 to simplify the design. In anticipation of the present publication, 

 therefore, I communicated to the Society a notice of a design to that 

 effect, which was read on the 9th December, 1880, and appears in the 

 "Proceedings " (vol. 31, pp. 141-146). 



Although, therefore, it precedes the present communication in that 

 respect, it will be understood to have grown out of the examination of 

 Broun's instrument, to which I am indebted for the means of judging 

 of its feasibility. 



I. Historical Review of Proposals or Designs for a Gravimeter. 



It is hardly necessary to describe at much length the object of 

 instruments of this class. They aim at determining statically that 

 which the differential pendulum has been had recourse to, ever since 

 the days of Graham and of Bouguer, to determine dynamically, the 

 variation of gravity at different times or places. The principle indeed 

 is so obvious that when we recognise the fact that hitherto nothing 

 has been effected which can be regarded as establishing a rivalry in 

 practice, we also perceive that the interest of any historical review 

 must centre in what it can teach as to the causes of the absence of any 

 practical competition. For this we must first learn what forms have 

 been proposed. 



It would be hard to say to whom should be ascribed the credit of 

 first suggesting the statical method. It would seem as if its theoretical 

 possibility must have long been perceived. Bat probably no clearer 

 enunciation of the principle is to be found than that which appeared in 

 the first, as in all later, editions of the " Outlines of Astronomy " 

 (Art. 234), first published in 1833.* So far as I know, what was there 

 proposed, divested as it purposely was of everything which could 

 confuse the principle to be illustrated, had never been experimentally 

 tried. To understand it, a single glance at the drawing, representing 

 a weight suspended by a coiled spring, its lower surface grazing the 

 base of the supporting frame, should suffice. The supposed variation 

 of gravity is intended to be balanced by the addition or removal of 

 small subsidiary weights, the proportion of which to the principal 

 weight will measure the relative variation. The only doubts which 

 arise at once in considering such an instrument have reference to the 

 means of exactly determining the contact and to the constancy of the 



* Art. 189 in the first four editions. 



