[ 2 6 2 ] 
are the fads, tho', perhaps, it may be hard to account 
for them from the common principles of motion. 
What feemed mod: likely to be the caufe of the verti- 
cal change of its direction, was its rapid defcent to- 
wards the earth, till it came fo low as that the refidance 
of the air might ad upon it ; in which cafe, the lower 
part of the body meeting with denfer air than the up- 
per, it would be fo refleded from that medium as to 
afcend ; for tho’, at the height of 20 or 30 miles, the 
air is extremely rarified, yet, in this indance, the im- 
menfe velocity of the impingent body will make a 
lefs redding medium produce a greater effed. The 
meteor, therefore, might be refleded by the air, in 
the fame manner as a cannon ball by water, when it 
drikes it in a very oblique diredion. 
If this reafoningcan be admitted, we may account 
for the lateral variation of its path, in the manner 
following. Although it fhould feem, that, during 
the fird part of the courfe, the wind was at S. E. 
yet, when the meteor advanced towards the north of 
Scotland, it there probably met with a wederly wind ; 
from this circumdance in Dr. Mackenzie’s letter, of 
the J'ea being that night remarkably louder than at any 
ether time during a whole month he had been at Flow- 
er dale : for that place lies on the wedern ocean, in 
the fldre of Rofs ; and there, I imagine, the fea could 
not produce that dormy noife, unlefs the wind had 
blown from fome point between the wed and the 
north. If therefore the current of the air obliquely 
oppofed the motion of the meteor, that body would 
be gradually refleded into a new diredion ; from 
which, by another current, it might have been turned 
a fecond or a third time, until it fell at lad into that 
path 
