ATMOSPHERIC REFRACTIONS. 167 
pareutly suspended in tlie air, at a considerable 
elevation above the actual termination of the 
mountains to which they refer. The whole exhi¬ 
bition is frequently a grand and interesting phan¬ 
tasmagoria. Scarcely is the appearance of any ob¬ 
ject fully examined and determined, before it 
changes into something else. It is, perhaps, alter¬ 
nately a castle, a cathedral, or an obelisk: then ex¬ 
panding and coalescing with the adjoining moun¬ 
tains, it unites the intermediate valleys, though 
they may be miles in width, by a bridge of a 
single arch of the most magnificent appearance. 
The cause of these phenomena, as far as they 
depend on refraction, is, I imagine, the rapid eva¬ 
poration which takes place in a hot sun, from the 
surface of the sea ; and the unequal density occa¬ 
sioned by partial condensations, when the moist air 
becomes chilled, by passing over considerable sur¬ 
faces of ice. The vapour produced by evaporation 
and partial condensation, is sometimes perceptible 
to the eye, rising like transparent steam in all direc¬ 
tions, in little curling clouds, and passing along 
with the breeze near the surface of the sea. Its in¬ 
fluence can sometimes be perceived at the distance 
of a few fathoms, or, perhaps, half a furlong, by the 
tremulous motion it appears to give to all bodies 
on the water or near it. In this case, it has a ten¬ 
dency to produce a serrated or basaltic appearance 
