[ + 2 7 ] 
to be the denfity of light upon the furface of the 
orbis magnus, and it will be found by computation, 
that its denfity upon the fun’s furface, mu ft be lefs 
than -i-^-j-d of the denfity of common water. 
From thefe confiderations, I think it may be con- 
cluded with the greatefl certainty, that the quantity 
of matter that ifiues from the fun in light, has been 
greatly over-rated in the foregoing computations. 
I apprehend, however, that the denfity of each 
feparate particle cannot be lefs than has been fup- 
pofed : but that the magnitude of each is lefs, and 
the arrangement lefs compact m . Let the denfity of 
each, and the number that ifiues at any one time 
from the fun, remain as before ; and let us confider, in 
what proportion the magnitude of each particle, mufi: 
be diminifhed, fo that they may altogether form a 
fluid, on the fun’s furface, 173 times lelsdenfe than 
water. Let A BCD reprefent a fe&ion of the fun’s 
fphere, through the centre S n . In the fpherical fur- 
face ABCD, take any point B. Join SB, and take 
Be , B f each equal to the femidiameter of a particle 
of light. Upon the centre S, at the diftances Sf, 
S /, imagine two other fpheres , eg h, fkl, one in- 
clofing, the other inclofed within the fphere ABCD. 
The folid fpace efklgh, is the fpace that contains 
all the particles of light (with their interftices) that 
iflue together from the fpherical furface ABCD; 
and becaufe e f bears an exceeding fmall proportion 
to S B, therefore the fpherical furfaces eg h , fkl, are 
very nearly equal each to the other, and to the fphe- 
m Vide Inftant. Produdh p. 37. “ See the fig. p. 429. 
I i i 2 rical 
