438 
MR. BAILY ON THE CORRECTION OF 
From these comparisons it appears that although, in the cases of the first 
four pendulums, the correction for the knife edge B exceeds that of the knife 
edge A, yet in the case of the thick brass bar (No. 31), the corrections for the 
two knife edges are nearly equal; and in the case of the brass tube (No. 35), 
the correction for the knife edge B is smaller than that for the knife edge A ; 
contrary to what takes place in the other pendulums, and contrary to the 
assumed theory on the subject *. 
M. Bessel has suggested, as one of the modes of rendering the two knife 
edges, of a convertible pendulum, synchronous, to make the figure symme¬ 
trical, but the mass not so : which may be effected by making one part of the 
pendulum hollow. In such case, however, we must consider the hollow portion 
of the pendulum as a substance of a different specific gravity, and compute its 
effect on the vibrating mass accordingly. The results also, in such a case, will 
probably differ according as the hollow portion is hermetically sealed, or com¬ 
municates freely with the circumambient air. 
■» 
Additional Experiments. 
Since this paper was read before the Society, I have made a number of addi¬ 
tional experiments on other pendulums of different forms and construction, 
and have varied and combined some of the preceding pendulums in several 
new modes ; with a view to clear up the anomalies apparent on the face of the 
preceding experiments, and to throw some light on the manner in which the 
air operates on the pendulum when in motion and affects the time of its vibra¬ 
tions. As the Council have given me permission to annex the substance of 
these experiments to the present paper, I shall briefly state the results obtained ; 
together with such explanation relative thereto, as may be requisite for under¬ 
standing the mode pursued, and the consequences deduced: but, I have not 
considered it necessary to encroach on this indulgence by giving the full de¬ 
tail of those experiments ; which, however, it may be proper to state have been 
conducted on the same principles and with the same regard to accuracy as 
those already given in this paper. Indeed, it will be seen that there is less 
* Probably tbe position of the two additional knife edges, with their knee-pieces, in the bar No. 31, 
and of the four additional planes, with their collars, in the tube No. 35, may have some influence in 
producing this apparent discordancy. 
