C 33 3 
after drefling) ; fo that the water was not quite ftill 
when they got into the ciftern : neither could I per- 
fuade all of them to lay themfelves down gently, 
much lefs to keep their heads under water fo long a 
time as one fecond : fo that, in moil of the obferva- 
tions, the furface of the water was far from being 
quite fo ftill, as to render the meafures perfectly ex- 
adt, I being obliged to catch them, as it were, by 
taking the mean height between the librations. More- 
over, the great area of the ciftern was no incon- 
fiderable bar to the accuracy I expelled. However, 
as I do not recoiled experiments of this kind any- 
where recorded, thefe, perhaps, may give fome fa- 
tisfadion to fuch perfons, who may have the curiofity 
to deftre fome knowlege on this fubjedt. Were I to 
make any more obfervations of this kind, I would 
chule an upright parallelopiped, not above 18 or 20 
inches in the lide of the fquare ; into which the per- 
fon fhould let himfelf down by fteps nailed to the 
lide : for in fo fmall an area the motion of the water 
would fooner fubfide ; neither would the librations 
be any thing near fo large as on a fmaller furface. 
One of the reafons, that induced me to make 
thefe experiments, was a deftre of knowing what 
quantity of ftr or oak timber would be fufticient to 
keep a man afloat in river or fea water, thinking 
that moft men were fpecifically heavier than river or 
common frefh water ; but the contrary appears from 
thefe trials : for, excepting the firft and laft, every 
man was lighter than his equal bulk of frefh water, 
and much more fo than his equal bulk of fea~water : 
conl'equently, could perfons, who fall into water, 
have prefence of mind enough to avoid the fright 
Vol. 50. F ufual 
