[ 92 ] 
Therefore the required probability, that the error, by 
taking the Mean of fix obfervations, exceeds not a 
fingle fecond, will be truly meafured by the fradion 
1088391168 ' " confequently the odds will be as 
78881480Q to 2p95’76368, or as to i, nearly. 
But the proportion, or odds, when one lingle ob- 
fervation is relied on, is only as 16 to 20, or as 
to. I.- — To find, now, the probability, that the re- 
fult comes within 2 feconds of the truth, let 
be made = — 2 ; fo fhall m (= — it') — — 12; and 
therefore And our 
general exprefTion will here come out 36079407 5 
and confequently £)— 10523 i J 761 : whence 
io883 ' 9mM meafure of the probabi- 
lity here fought ; and the odds, or proportion of 
the chances, will therefore be as i05'2 3U76i to 
36079407, or as 29 to I, nearly. But the propor- 
tion, or odds, when one fingle ohfervation is relied 
on, is only as 2 to i : fo that the chance, for an 
error exceeding 2 feconds, is not tV part fo great 
from the Mean of fix, as from one fingle obferva- 
tion. And it will be found, in the fame manner, 
that the chance for an error exceeding 3 feconds, will 
not be To'o~ part fo great from the Mean of fix, as 
from one fingle ohfervation. Upon the whole of 
of which it appears, that the taking of the Mean of 
a number of bbfervation-s, greatly dirninifhes the 
chances for all the fmaller errors, and cuts off al- 
mofl: all poffibility of any great ones : which laff 
conlideration, alone, fcems fufiieient to recommend 
? the 
