a 
432 Different “ Rates” of Pennington’s Astronomical Clock, 
Barometer. | Fah. Thermometer. 
Day 1817. Nese a 
Max. Min. || Max. | Min. |-Diff. 
July 31 || :29°411 | 29°223 | 6s° 54° 14° 
August 1 || 29333 | 29315 | 61 50 |. 11 
2 || 29°434 | 29413 | 57 49 - 8 
3 || 29530 | 29°'422 | 68, 53. 15 
| 4 || 29488 | 29426 || 50 Si a 
| 5 || '29°:737 | 29°663 51 AG alpi-3 
6 || 29°707 | 29°687 504 | 50 4 
7 || 29°646 | 29°632 534 | 52 14 
8 || 29°641 | 29°638 || 554 | 521 3 
Q || 29°752 | 29°730 57 52 5 
10 || 29°834 | 29°826 | 61 54 4 
11 |) ee we | ee ee 64 53 11 
ase i4 eee 634 | 55 8} 
13 || 29°892 | 29°864 62 53 9 
14 || 29:903 | 29942 664 | 53 134 
15 || 29.941 | 29'920 59 54 5 
TO | et ei vies 56 52 4 
The thermometer by which the temperature was measured, 
hung upon the side of the clock-case; the barometer which served 
to measure the atmospheric pressure, hung within two feet of 
the clock. The great differences between the maximum and 
minimum temperatures on July 31st, and August Ist, 3d, 1] th, 
and 14th, were occasioned by the sun’s rays striking upon the 
slightly elevated tarpaulin roof of the transit-house, thus greatly 
raising the afternoon temperature, and by the wide opening from 
the top to the bottom of the roof during observation, which gave 
a corresponding depression to the temperature at night. 
The rapid depression of from 13° to 15° of temperature, took 
place on three afternoons between four o’clock and ten: yet this 
produced no appreciable change in the rate: from which we may 
infer, that the compensation in the pendulum is exceedingly per- 
fect, and that the instrument does not yield to the changes of 
temperature “ by starts,” as has been often objected to the grid- 
iron compensation. 
Again; the ‘‘ rate” of the clock, not varying from its mean 
rate more than one-fifth of a second on any day during the ob- 
servations, although the barometric column varied from 29:22 
to 29:96; we may, I think, fairly infer, that if the barometric 
changes do not exceed an inch, in the course of a series of com- 
parative experiments with this clock, the introduction of any al- 
lowance for the air’s buoyancy or its resistance, would be a use- 
less refinement in the computation. 
We 
