4jo Place of an Olje ft feen through a Window. [Book HI. 



the preceding rules, is xr> and let Qinop reprefent 

 the progrefs of one of the rays diverging from Q^ 

 Then, from what has been before obferved, xp is pa-^ 

 rallel to Qni. Therefore q m : m o : : Q^ : Qj, that 

 is, when m is very near to E, 



qE : EF :: ,Q^ : Q# } 

 or, ^E: Q^q :: EF : Qjc. 



But fuppofing I : R to reprefent the ratio of the fines 

 of incidence and refra6tion of a ray pafling into the 

 glafs, 



?E: Qj>::I : I R, 

 .-.EF:Q* ::I:I R; 



that is, the interval between the furfaces of the pane is 

 to the diftance between the real and apparent places 

 of the object as the fine of incidence to the difference 

 between the fines of incidence and refraction. For 

 glafs this ratio- is nearly that of 3 to a ; therefore I : I 

 R : : 3 : i, and Q^ will be therefore one third of 

 E F ; if the pane of glafs is a tenth of an inch thick, 

 an object fren through it will not appear to be a thir- 

 tieth part of an inch out of its real place ; a change 

 which is too fmall to be taken notice of in Common 

 life. 



A ray paffing through a medium bound by plane 

 furfices inclined to each other, is bent towards the 

 thicker or thinner part of the medium, according as 

 the medium is denfer or rarer than that by which it is 

 furrounded, and the place in which an object will ap- 

 pear to be is found by a very eafy conftruction. Let 

 ABC (Plate XV. Fig. 16.) be a glafs prifm, QJR an 

 object feen through it by an obferver at O, From Q^ 

 draw QJi perpendicular to the firft fiirface A C, and 

 let q be the focus of rays refracted by that furface ; 

 fr-.>m q draw <?E perpendicular on AB, the frcond fur- 

 face produced, if neceflary, and fuppofing q then to be 



