146 On a Simplified Form of Tors ion- Gravimeter. [Dec. 9, 



Finally, the desired result will depend on the weights. I consider 

 that to weigh accnrately even a large body of water is a thing that can 

 certainly be done, but how to do it forms no part of my present design. 

 The drum might conveniently be constructed with a false bottom, the 

 real bottom of the water space being conical, with the stopcock at the 

 apex. 



By an easy arrangement the closing of the cock might be effected 

 as a consequence of the critical point of unstable equilibrium being 

 reached and passed, the swing being checked by a stop pressing on the 

 cock and shutting it off. Being thus self-acting, the flow might be 

 made as slow as desired, with the certainty of the proper result being 

 reached without further attention. 



The quantity of water remaining could then be measured at leisure. 



The change of gravity which would cause a change of rate of one 

 second per diem in a seconds pendulum is 4 3 1 00 part. On a weight 

 of 200 lbs., or 1,400,000 grs., this would be 32*4 grs., or rather more 

 than half a fluid dram. 



I am, of course, quite aware that the efficiency of such a gravimeter 

 depends ultimately on the constancy of the torsion. This is a sine qua 

 nun in any gravimeter on the torsion principle, and I assume it to 

 exist. If it does nob exist, no gravimeter on this principle can be 

 successful. 



Nor am I regardless that this, and all other gravimeters depending 

 on elasticity, are at the mercy of temperature. But however impossible 

 it may be to command a definite and uniform temperature wherever 

 such an instrument may be instated for observation of the change of 

 gravity, the converse is possible enough : the constancy of gravity can 

 be guaranteed* while observations are instituted for determining the 

 effect of thermometric change. So that if insufficient knowledge of the 

 variation of elasticity with temperature debars such an instrument 

 from being used as a gravimeter, it may with the more reason be looked 

 to to increase that knowledge. 



It seems only necessary to add that the mode of utilising the fact of 

 a position of unstable equilibrium existing, if desired, betweeu 0=90° 

 and 0=180°, as here indicated, may be open to objection without pre- 

 venting that fact being capable of utilisation in some better way. The 

 need of a statical gravimeter is so great that I hope what I have said 

 may draw attention to the subject by showing how very simple it really 

 is. Under correction, I would hazard the assertion that there is no 

 instrument which would command a more immediate field of useful- 

 ness than a simply constructed gravimeter, and I think I have shown 

 that it is possible to construct one. 



# The contrary at any rate is, at the most, suspected more or less strongly, by a 



few. 



