Ixxxvi 
APPENDIX. 
horizon, and when altitudes, or zenith distances of the stars, were required for 
the purpose, the adjustments and management of the circle being necessarily 
performed by lamp-light, could not be got through with sufficient expedition, 
combined with the care requisite to ensure accuracy, before the completion 
of the observation was stopped by the causes above-mentioned. 
As the employment of the repeating circle had been looked forward to for 
various purposes during the winter, and especially for observations on the 
amount of atmospherical refraction in extreme cold, its failure caused much 
disappointment ; so long, however, as the temperature remained above the 
freezing point of mercury, the error of the chronometers on time could be 
obtained by altitudes with a sextant and artificial horizon with much facility and 
with tolerable accuracy ; the mercurial horizon was used even so low as —41°, 
by previously warming the mercury and the leaden trough into which it was 
poured, and by completing the observation before they had time to cool. 
