Prifmes 5 as is evident by the Tangents A D> drawn to the 
Grain A, at the points , where the Ray D A enters or corns 
out. 
Further he took notice, that it was alfo manifeft , why the 
r^/s? colour is in the interiour circumference of the Halo. And 
whythcfpace, which it taketh in, and chiefly near the moft 
lively colour'd parts , appears obfcurer than the Air about-, viz. 
becaufe 'tis there, where Grains are , which tranfmit no 
xays of the Sun to our Eyes , and fodo nothing but darken 
the Air, as the drops of water when it raineth. 
Moreover he noted , that M, Des Cartes ^ endeavouring to 
explain the Caufe of th^k Halo's^ had committed amiftake, 
for want of Obfervations , truly relating this laft circum- 
ftance: Becaufehe maintains, that the fpace comprifcd with^ 
in the Halo is clearer than the Air withoutj-^ndto renders 
reafon of it, he fuppofeth certain Grains, tranfparenr, 
having the • form of a lentil 5 which fuppofition cannot there- 
fore be true , becaufe what he deduceth from it , is contrary to 
whatisobferved: Befides that the roundnefs of the ^4/^ in all 
the elevations of the Sun agreethnot with it, as wereeafyto 
flaow. 
He ftayed not to examine the Qeneration of the half-dark 
grains , becaufe he is (liortly to publifli a larger Treatife of this 
fubjecS^ wherehe intends alfo to render a reafon, why the 
dark kernel keeps certain proportions, rather than others^with 
the Grain, which contains it. 
A% to the Arch of the Circle , which above ,touched the 
Halo (ttn May 12th, as alfo that the C:)lours were 
more vivid in this place , and inxhat below, then in the reft of 
the Circle 5 he faid^that thefeefFedls did not proceed from the 
G ains, he hadbecn fpeaking of, but from another caufe, which 
did alfo ferve for the produ(ftionof the Parheliaznd the Circles 
which almoft always accompany them, 
1 ouching which Circles and Parhelia s he told the Company, 
thatbtfides the round and half-dark Grains, there were alfo 
formedin the Air certain little Cylinders of the like nature, and 
of which M. DiTxC^rtohimfelf declared in his Treatife if Me- 
tem tohaveobfcrved fome, not indeed with opaque kernels 
within. 
