1881.] 



Major J. Herschel. On Gravimeters. 



319 



If a heavy bar be suspended by two parallel inelastic lines, the force 

 which is exerted when the position of rest has been quitted varies, in 

 all positions, as the force of gravity at the time : if the angle of dis- 

 turbance and all other attendant circumstances be strictly the same, 

 the proportion between the vertical and horizontal forces is constant ; 

 and if the suspending lines are long, as compared with their distance 

 apart, this constant is a small fraction. This is " le fractionnement 

 de la pesanteur " of which Babinet speaks. We are to regard the 

 horizontal force as representing gravity and varying with it, at 

 different times or places. Now, if the suspension had been by a single 

 instead of a double line, and that elastic, the horizontal force exerted 

 in consequence of disturbance would have been independent of the 

 weight of the bar, and therefore also of variation of gravity. It 

 would be a constant force, as far as gravity is concerned. 



In the torsion gravimeter these two kinds of suspension are com- 

 bined — the constant being adjusted so as to balance exactly the in- 

 constant force, by the addition or subtraction of a minute subsidiary 

 weight to the primary weight carried by the double lines. 



It would be difficult to explain briefly the mechanism by which the 

 condition of equilibrium is ascertained to be perfectly secured : 

 enough has been said to render the source of failure intelligible. 



Success must evidently depend on constancy in the force exerted by 

 the single wire ; and this on the fixity of attachment of the ends of 

 the same. It appeared by repeated trials and experiments of different 

 kinds that this attachment was faulty : further experiments, having 

 for their object to remedy the defect, proved unavailing ; and at last 

 the attempt — and with it the intended trial of the instrument as a 

 gravimeter — was abandoned, as under the circumstances hopeless. 



Unsatisfactory as this result undoubtedly is — especially for the 

 credit of so developed an instrument — the inquiry has not been 

 entirely barren of useful result. The theory of the machine had not 

 been fully considered, and at least one important point would seem to 

 have escaped recognition altogether. There is reason to think that 

 had it been seen that the double suspending lines must practically 

 perform a function precisely similar to that for which the single wire 

 is provided, the latter would have been dispensed with. It is true 

 that this implies not only other modifications of design but also the 

 further recognition of the part actually played by the torsion of the 

 double wires, of which no sign appears in anything that has yet been 

 published about it. But it is scarcely doing violence to the ingenuity 

 which developed a complicated machine to suppose that it would have 

 seized upon and utilised a mode of simplification had attention been 

 directed to it. 



The simplification to which allusion is here made, has been brought 

 to the notice of the Society in a previous paper. 



