J 4^ Phtntmina of unufual Refrafl'ion> 



defcribed. For let b (fig. lo), be the furface of the fea ; a b z, an objefV ; E, the place of the 

 eye ; a r E, ^ J E, the progrcfs of two rays by the ufual rcfr?i£tion from tlie extreme parts of 

 the objedt to the eye; to thefe curves, draw the tangents E a\ E b'-, and a' b' will be the image 

 of the obje£t, as ufually formed. Now if we take the cafe reprefented in fig. 4. let a' b" 

 reprefent the inverted image, and a" b'" the ereft image. Join a E, a" E, and b" E, b" E, 

 and thefe lines muft refpedtively be the diredlions of the rays entering the eye, from a b, in 

 order to produce the images a b'\ and a" b', hence thefe lines muft be tangents at E, to 

 the curves which are defcribed by the rays of light; let, therefore, a n E,, a m ¥.-, b v E, b tv Ey 

 be the curves defcribed. We have, therefore, to affign a caufe which may bring rays pafling above 

 the rays a r E, b sE,to the eye at E. Now if there were no variation of the refraftive power 

 of the air, a ray of light paffing through it would defcribe a ftraight line; therefore the curvature 

 of a ray of light paffing through the atmofphere depends upon the variation of the refractive 

 power of the air. If, therefore, we fuppofe the air lying above a r E to vary quicker in its 

 jefradive power than the air through which a r E pafles, the curvature of a ray pro- 

 ceeding atove that of a r E will be greater than the curvature of « r E ; and upon this 

 principle we may conceive that a ray may defcribe the curve a n E, and in like manner, if 

 a quicker variation of refrafliive power (hould take place above the curve a n E, thin in that 

 curve, a third ray may defcribe the curve a mE. The fame may be faid for the rays b v E, 

 b wE, diverging from b. The alterations of the refradtive power may arife partly from the 

 variations of its denfity, and partly from the variations of its moifture ; and the paflage of the 

 rays through the boundary of the fog, may there fufFer a very confiderable refradtion; for from 

 the motion of the fog, and that of the images above mentioned, I have no doubt that the fog 

 was a very confiderable agent in producing the phenomena. When all the caufes co-operate, 

 I can eafily conceive that they may produce the efFedts which I have defcribed. If the caufe 

 fhould not operate in the traiSt of air through vvhich the curves a n E, b v E pafs, but fhould 

 operate in the trail, through which a mE, bw E pafs, an erefl: image which would be vifible, 

 but there would be no inverted image; and {hould it operate in the latter cafe, but not in the 

 former, there would be only an inverted image. 



As the phenomena are very curious and extraordinary in their nature, and have not, that I 

 know of, been before obferved, 1 have thought jToper to lay a defcription of them, with all the 

 attending circumftances, before the Royal Society. They appear to be of confiderable im- 

 portance ; as they lead us to a knowledge of thofe changes, to which the lower parts of the 

 atmofphere are fometimes fubje6t. If when thefe phenomena appear, a veffel furniihed with a 

 barometer, thermometer, and hygrometer, below, and alfo at the top of the mart, were fent out 

 to pafs below the horizon, and return again, and an obferver at land, having like inftruments, 

 were to note, at certain intervals, the fituation and figure of the images, it might throw further 

 light upon this fubjeft, and lead to ufeful difcoveries refpeftiiig the ftatc of the atmofphere, from 

 aconjundtion of the caufes which a iFedl thefe inftruments. 



Jtt 



