VOL. XXI.] PHILOSOPHICAL TRANSACTIONS. 433 



come to the cement, or insinuate itself through its pores or fissures, if any 

 happened to be left unclosed. With all these precautions, the experiment suc- 

 ceeded at last as wished, after this manner. 



We placed the object a, which was a black thread sliding in a little frame over 

 a piece of white paper, in the axis of the cylinder ex continued to it, we filled 

 the pipes and cylinder with mercury, and having stopped the upperniost pipe at 

 p with the small iron stopple k, and closed it at the other joints, we let the 

 mercury run out gently at o into the bladder u, till it remained suspended at the 

 usual height, as in the barometer, leaving the space above it between the glasses 

 oo and ff void of air. We then found the object, which before appeared in the 

 axis at x, raised considerably above it, and we reduced it to appear at x by removing 

 it from a to «. The axis therefore of the visual ray xa, which was also the axis 

 of the cylinder xca, falling perpendicularly on the void space in the cylinder, 

 passed through it without any refraction ; but emerging obliquely into the air, 

 it was refracted towards the perpendicular pc, and there received a new direction 

 to a.. And therefore the space a « subtended the angle of refraction acx, which 

 we measured, and found as follows : 

 The height of the object above the axis of the unrefracted visual inc. dec. 



ray a C6 0,425 



Thedistanceof the object from the refracting plane ac about 5lfeetor 0,612 



Therefore the angle of refraction aca was 0° l' 23* 



The angle of emersion pea (by the construction of the cylinder) was 62 30 O 



Therefore the angle of incidence pcai := (= pea + aca) was .... 62 17 37 



And therefore universally, according to the known laws of refraction. 



The sines of the angles of incidence being 100000 



The sines of the angles of emersion are 100036 



And the refractive power of the dense air 36 



By the refractive power of a pellucid body, I mean that property in it by 

 which the oblique rays of light are diverted from their direct course, and which 

 is measured by the proportional differences, always observed, between the sines 

 of the angles of incidence and emersion. 



This property is not always proportional to the density, at least not to the 

 gravity, of the refracting medium. For the refractive power of glass to that 

 of water is as 55 to 34, whereas its gravity is as 87 to 34, that is, the squares 

 of their refractive powers are very nearly as their respective gravities. And 

 there are some fluids which, though lighter than water, yet have a greater power 

 of refraction ; thus the refractive power of spirit of wine, according to Dr. 

 Hook's experiments, Microg. p. 220, is to that of water as 36 to 33, and its 

 gravity reciprocally as 33 to 36 or 36'-. But the refractive powers of air and 



VOL. IV. 3 K 



