VOL. XI.J PHILOSOPHICAL TRANSACTIONS. 2/9 



long image into a round one, by the contrary refraction of a second prism, 

 further explained in N° 83. For this is not to be done in Mr. Linus*s surmise 

 of a parallel image, and therefore had Mr. Linus considered it, he could never 

 have run into that surmise. 



This I suppose is enough to manifest the three particulars; any one of which 

 being evidenced, is sufficient to take away the scnaple. And therefore Mr. 

 Linus's friends need not fear but that the further directions I sent them lately 

 for trying the experiment, are the same with those I have followed from the 

 beginning ; nor trouble themselves about any thing but to try the experiment 

 right. But yet, because Mr. Gascoin has been pleased to insinuate his suspicion, 

 that I do differ from himself in those directions, I shall not scruple here to re- 

 duce them into particulars, and show where each particular is to be found. 



1. Then he is to get a prism with an angle about 6o or 65 degrees, N^ 80. 

 If the angle be about 63 degrees, as that was which I made use of, N° 80, he 

 will find all things succeed exactly as I described them there. But if it be larger 

 or less, as 30, 40, 50. or 70 degrees, the refraction will be accordingly larger or 

 less, and consequently the image longer or shorter. If his prism be pretty 

 nearly equilateral, such as 1 suppose to be usually sold in other places as well as in 

 England, he may make use of the largest angle. But he must be sure to place 

 the prism so, that the refraction be made by the two planes which comprehend 

 this angle. I could almost suspect, by considering some circumstances in Mr. 

 Lmus's letter, that his error was in this point, he expecting the image should 

 become as long by a little refraction as by a great one; which yet being too 

 gross an error to be suspected of any optician, I say nothing of it, but only hint 

 this to Mr. Gascoin, that he may exam inc. all things. 



2. Having such a prism, he must place it so, that its axis be perpendicular to 

 the rays N° 84. A little error in this point makes no sensible variation of the 

 effect. 



3. The prism must be so placed, that the refractions on both sides be equal, 

 N* 80 : which, how it was to be readily done by turning it about its axis, and 

 staying it when you see the image rest between two contrary motions, as I ex- 

 plained in my late descriptions, so I hinted before, N° 80. If there should be a 

 little error in this point also, it can do no hurt. 



4. The diameter of the hole 1 put -i- of an inch, N° 80, and placed the prism 

 close to it, even so close as to be contiguous, N*^ 80. But yet there needs no 

 curiosity in these circumstances. The hole may be of any other size, and 

 the prism at a distance from the hole, provided things be so ordered, that the 

 light appear of a round form, if intercepted perpendicularly at its coming out of 

 the prism. Nor needs there any curiosity in the day. The clearer it is the better; 



