Theory of Horizontal RefraBio't, 149 



hrcreafing the diftance) the two images would appear continually to approach each other, 

 till blended with double light in one, and difappear at the altitude ;, above the apparent 

 horizon of the fea. But, as explained before, if the ftrength of evaporation did not fe- 

 parate by refraction the pencils a a and d d to a greater angle than double the angle that the 

 lamps and reflettors appear under, the two images would be blended, and the ftrong ap- 

 pearance of light would be of fliorter duration. The diftance run from the lights during the 

 time each of the lights (hone bright, would have been ufeful ; but this did not occur at the 

 time, nor have I had the like opportunity fince. However, I recommend to the mariner to 

 ftation people at different heights in looking out for a light, in order to get fight of it near the 

 horizon, when it is always ftrongefl. 



Refpefting the appearance of the Abbey-head,'before mentioned. Fig. i, the dotted line 

 AB reprefents the limit, or the loweft points of the land ihat can be feen over the fea ; for, as 

 above Rated, all the objeiSs appearing below this line are the land above it inverted ; and 

 where the land is low, as at d and m, it muft appear elevated above the horizon of the fea. 



In Figure 5, let HO reprefent the curve of the ocean, and d rhe extreme top of the mount 

 vifible at A by the help of refra£tion ; the dotted pencil of rays cc pafTing from d to the eye, 

 in fome part a little below the maximum of denfity, where inverfion begins; therefore no 

 land lower than this can be feen ; for any pencil from a point in the land lower than 

 this muft, in the refraftion, have a contrary flexure in the curve, and therefore pafs above 

 the obferver. Let AD be a tangent to the curve at A, then the object d will appear to be 

 elevated by refraftion to D ; alfo let A v be a tangent to the pencil A x at A, then the 

 angle DA a' will appear to be an open fpace, or between D and the horizon of the fea. 

 Suppofe a ftar ftiould appear very near and over the mount d, as at *, two pencils would 

 iflue from every point of it, and form a ftar below as well as above the hummock d. There 

 are always confufed or ill defined images of the objefts at the height of the dotted line. 

 Fig. I, above the level of the fea, as before mentioned ; and inftead of the points of d end- 

 ing fharp in that line, they appear blunted, and the Abbey-head is frequently infulated at 

 the neck /«. 



I have viewed from an elevated fituation a point or headland at a diftance beyond the 

 horizon of the fea, forming, as in Fig. 6, a ftraight line AB, making an acute angle BAO 

 with the horizon of the fea. Seeing the extreme point blunted and elevated, I defcended ; 

 and though in defcending the horizon cut the land higher, as at HO, HO, yet the point 

 had always the fame appearance ss naa, Fig. 6, though the land is known to continue in 

 the dire£lion of the ftraight line AB to beneath the horizon, or nearly fo, as viewed from 

 the height above. 



K then, from a low fituation, we view this headland through a telefcope, the inclination 

 of the furface A B to the horizon being known to be a ftraight line, it will appear as ia 

 Fig. 7; the dotted line (at the height of the point where a perpendicular .vji would touch the 

 extreme of the land) being at the limit or loweft point of cre£l vifion. And if a tangent 

 to the curved appearance of the land n i is drawn parallel to the inclined furface of the 

 land A B, Fig. 6, touching it at C, the point C will Ihcw the lieight of the maximum of 

 dcnfity, where the pencil of the rays of light from thence to the eye approaches ncareft the 

 fea; for pencils of rays from this land, taken at fmall diftances from C, will form parallel 

 curves nearly through the refracting mediums, and C will be the point of grcatclt refrac- 



ti6n } 



