338 CAPTAIN SABINE ON THE REDUCTION OF CAPTAIN RATER’S PENDULUM. 
computed on the simple consideration of buoyancy : they were particularly so 
in those instances in which the extremities of the pendulum were of the lighter 
material ; for the increase in the retardation in those instances much exceeded 
the proportion due to the diminution of the general specific gravity of the pen- 
dulum, occasioned by the addition of the small portions of wood. 
In viewing the comparative retardations in the two positions of the pendulum, 
in each of these experiments, we find a confirmation of the inference, noticed 
in the earlier part of this paper, that in consequence of the want of symmetry 
in the two ends, the reduction to a vacuum ought not to be the same in the 
two positions, of the pendulum. We have also a curious exemplification of 
the influence of the addition of equal portions of matter at each extremity of 
the pendulum, in diminishing the difference in the retardation occasioned by 
the disparity in the form and size of the weights. With the short wooden 
tail pieces, the difference (which with no tail pieces at all would probably have 
exceeded 3 vibrations per diem) amounted to 2.5 vibrations. With the brass 
tail pieces it was lessened to 1 vibration. And with the wooden tail pieces of 
their original length, the effect of the inequality of the weights was almost 
altogether counterbalanced, the retardation being within half a vibration the 
same in each position of the pendulum. 
Finally, it is curious to perceive how much the result obtained by Captain 
Kater, as the length of the seconds pendulum, depended on the mere accidental 
circumstance of the addition of tail pieces to his experimental pendulum : had 
the circumstances of the experiment been varied in regard to the tail pieces ; 
had they been of brass for example ; — or being of wood, had they been of any 
other length than that which was determined by the accidental circumstance 
of the relative heights of the clock, and pendulum support ; — or had they been 
altogether omitted and the coincidences observed by means of the bar itself, — 
a widely different result would in each of these cases have been arrived at. 
