470 
MR. BAILY ON THE CORRECTION OF 
on the invariable pendulum made by the English, the initial arc has been 
about 1°*2 or 1 0, 3 : but in those made by MM. Freycinet and Duperrey, the 
initial arc has sometimes amounted to upwards of degrees; and Mr. Rum- 
ker, in his experiments on the length of the seconds pendulum, has, in one 
instance, commenced with an arc of 11 degrees *. 
On Captain Sabine’s recent determination of the length of the seconds pendulum 
at Greenwich. 
In the volume of the Philosophical Transactions just quoted. Captain Sabine 
has also given, what he considers, the true length of the seconds pendulum at 
Greenwich ; and which he makes equal to 39T3734 inches, as deduced from 
his own observations there. It is not my intention to make any remark on 
those observations; which, indeed, appear to have been made with all due 
regard to accuracy: but, I trust I may be allowed, whilst treating on a sub¬ 
ject of this kind, to express my dissent from the inode in which he has deduced 
the result in question. In all cases of the convertible pendulum, either the 
perfect synchronism of the two knife edges, or (which will answer the same 
purpose), the difference in the results of the two knife edges, ought to be well 
established, by an equal weight of evidence for each knife edge. This is indis¬ 
pensable : and, unless it be accomplished, the problem cannot be considered 
as strictly solved. Each knife edge is independent of the other; and each 
ought to have equal weight in the determination of the result. It is true that 
the knife edge A (or that position of the pendulum where the great weight is 
below the axis of suspension), will, in case of any difference, always give a 
result nearer to the true value than the knife edge B : but, the proper cor¬ 
rection to be applied to make them synchronous, can only be determined by 
first giving to B an equal weight in the experiments. Now, perfect synchro¬ 
nism I consider unattainable ; or, at all events, not worth the trouble it would 
cost to pursue it: since the small difference which arises, in these cases, will 
always enable us to apply the proper correction, from the known principles of 
the pendulum ; and which are a more sure guide on such occasions than any 
partial determination of the correction from actual experiment, where, in these 
minute inquiries, the errors of observation are sure to baffle us in our object. 
* Memoirs of the Astronomical Society, vol. iii. page 289. 
