” 
444 Different “ Rates” of Penningion’s Astronomical Clock. 
from such operations, appropriate corrections for the changes in 
temperature and in atmospheric pressure, are applied to two mea- 
surements; viz. that of the diameter of the ball of the pendulum, 
and that of the vertical distance between the plane of suspension 
and the horizontal metallic plate which is screwed up to become 
a tangential plane to the ball of the pendulum when hanging at 
rest. Of these measurements, I venture, with all proper defe- 
rence to the ingenuity of M, Biot, to question the accuracy of 
the second; and itis due to the cause of scientific truth to state 
why I entertain doubts. The adjusting screw, &c. by which the 
metallic plate was elevated to come into contact with the bottom 
of the ball, were fixed upon a prismatic block of stone which, 
both at Leith and in Unst, was embedded in the earth. The 
consequence was, that when M. Biot, or any other individual 
whom he permitted to approach, stood within a foot or two of 
the apparatus, as he set down or took up one of his feet, the 
stone with the adjusting screw, metallic plate, &c. perceptibly 
became elevated or depressed, so that in one case the ball would 
touch the plate, in the other vibrate freely. Several persons 
besides myself witnessed this effect, both at Leith Fort, and in 
Mr. Edmondston’s cow-house where the apparatus was placed i in 
Unst: M. Biot, indeed, was in the habit of advancing and re- 
ceding within suitable limits, to show to strangers this motion in 
the stone and what it carried. Now, as it was not possible to 
bring up the metallic plate to touch the ball, and to measure the 
distance between that plate and the plane of suspension, without 
the observer’s moving to and fro, and at the same time moving 
the stone, the question naturally arises, How was error preclu- 
ded ? If it be replied, that the conductor of the experiment al- 
ways knelt down gently on a certain point when he screwed up 
the plate, and stood on that point, or some other constant point, 
every time he approached (for he must approach several times, 
see Biot’s Physical Astronomy, vol. iii. p. 159. Additions) for 
the purpose of applying his metallic ruler to measure the above- 
mentioned distance; the obvious remark is, that notwithstanding 
those precautions there would probably be a constant deviation 
of the horizontal metallic plate from its true position ; and that 
since the change of position in the plate was perceptible during 
different positions of the observer, it is further probable that the 
said constant deviation, should it be such, far exceeds the allow- 
able limits of error in an experiment which pretends to an accu- 
racy expressed by a recorded result carried to the ninth place of 
decimals*. Is this deviation appreciable? If it be—upon what 
principles, and by what process ? 
* M. Biot gives *994948151 of a metre as the length of the second pen- 
dulum at Unst. He does not specify to what temperature this determination 
ris eferred. Nothing 
