< 701 ) 
San might appear furrounded with darknefs, or if the Suns 
light were much ftronger chan it is, the colours would flill 
appear ro a greater length. 
In all thefe Obfervations the breadth of the Image was juft 
two inches. But obferving, that the fides of the two Prifms, I 
ufed, were not exa61y plait), but a little convex^ (the convexity 
being about fo mucli as chat of a double Convex-glafs of a 
fixteen or eighteen foot f^/fp^/»^) I took a third Prifm, whofe 
fides were as much coocave as thofe of the other were convex ; 
and this made the breadth of -^e Image co be two inches and a 
third pare of an inch ; the angles of this Prifm, and the 
lengths of the Image made by each of thofe Angles being thofe 
expreft in this Table. 
Ihe Angles of theFrifm. 
degr. 
58 
591 
62i 
the Lengths of the 
Image in inches. 
9 
io4 . 
In this cafe you fee, the concave ^figure of the fides of the 
Prifm by making the rays diverge a little, caufes the breadth of 
the Image to be greater in proportion to its length than ic 
would be otherwise. And this I thought fit to give you no- 
tice of, that IAt.Lhcos may examine^^whether his Prifm have not 
this fault. If a Prifm may be had with fides exaflly plain 5 ic 
may do well to try the Experiment with that 5 but its better, 
if thefidesbeaboutfo much convex as thofe of mine are, be- 
caufe the Image wjlT thereby become much better defined. 
For this convexiry^of the fides does the fame effeO, as if you 
fliould ufea Prifm with fides exaftly plain, and between it 
and the hole in the Window-fliut, placeanObjeS-glafsof an 
foot Tele fcope^ to make the round Image of the Sun appear 
diftinftly defined on the wall when the Prifin is taken away, 
and confequently the long Image made by the Prifm to be much 
more diftinftly defined (efpecially at its fireight fides) than it 
would be otherwife. 
One thing more I fliall add : That the utmoft length of the 
Image from the fainteft Red at one end to the fainteft Blew at 
the 
