SOUTHERN AFRICA. 107 
ciently ftrong to determine ; but a variety of circumftances 
feem to favor the former fuppofition. 
In looking through the exhalations of thefe beds of nitre, a 
meteorological phenomenon ^ of a different nature, was alfo here 
accidentally obferved. In marking about funrife the bearing 
by a compafs of a cone-fhaped hill that was confiderably ele- 
vated above the horizon, a peafant well acquainted with the 
country obferved that it muft either be a new hill, or that the 
only one which ftood in that diredion, at the diftance of a 
long day's journey, muft have greatly increafed of late its di- 
menfions. Being direded to turn his eyes from time to time 
towards the quarter on which it ftood, he perceived, with 
amazement, that, as the day advanced, the hill gradually funk 
towards the horizon, and at length totally difappeared. The 
errors of fight, occafioned by the refradive power of the air, 
are fo fmgular, and fometimes fo very extraordinary, as 
hitherto to have precluded the application of any general 
theorem for their correction, as it is not yet afcertained even 
through what medium rays of light, in their paflage, fuffer 
the greateft and leaft degree of refradlion. Were this precifely 
known, obfervations on the fubjed; might lead to a more inti- 
mate knowledge of the nature of the different currents of air 
that float in the atmofphere, and without doubt are the caufe 
of extraordinary appearances of objeds viewed through them, 
A gentleman, to whom the world is much indebted for his 
many ingenious and ufeful inventions and difcoverles, once 
propofed to determine the refradive power of different liquids 
and aeriform fluids ; and it is to be hoped he ftill means to 
P 2 profecute 
