THUNDER STORMS, 99 
are firft colle&ted into a body and condenfed, and thereby 
rendered vifible, which in a thunder cloud will be in the 
weft, notwithftanding the vapors of which it confifts may 
chiefly have arifen from the fea. 4. That when a thunder 
cloud is once formed it will be ina ftate of attraCtion with 
the earth in general, and more efpecially fo with all fub- 
ftances which are natural conductors of the eleCtric fluid, 
fuch as the water contained in rivers, bays, arms of the 
fea, &c. and by thefe the courfe of a thunder cloud is 
known to be very fentibly affected. 
But the ocean is the grand obje& towards which its 
courfe will be direted; accordingly the progrefs of the 
clouds is from the weftern horizon, eaftward, be the wea~ 
ther below what it may, not excepting the moft violent 
eafterly {torras, which are fometimes, though but rarely, 
accompanied with thunder and lightening. 
To the foregoing obfervations I would add, 5. That 
when an extenfive thunder cloud is forming in the atmof- 
phere by means of the mutual attraction of the conden- 
ding vapors, and the body of ele&trifed air which fuftains 
and condenfes them, the increafing denfity of the whole 
compound mafs of air and vapor will, by degrees, occafion 
its redefcent towards the earth, from the law of gravity; 
it will alfo be attracted by, and move towards the ocean, 
upon the principles of electricity; the cloud will then de- 
fcend obliquely, in a diagonal between the direGtions of 
thefe two powers; and both, continually acting upon it, 
will jointly accelerate its motion. Such a cloud, if denfe 
and large, would end in a perfect tornado, either upon the 
land or water, as thunder fhowers frequently do; fmaller 
clouds being alfo, ufually, accompanied with "eufts or 
flurries of wind. 
I fhall here add one obfervation more which I have fre- 
quently made, and which may tend to confirm the fore- 
going theory, viz. That as the general courfe of the eaftern 
coaft "of north America is from north-ea{t to fouth-welt; 
N 2 the 
