5JJ liUANTITY OP HORIZONTAL KEFRACTION. 



The object that I have at all times chofen, as fiiewing bed 

 the quantity of refraction, has been either an oar dipped in the 

 water at the greateft difcernible diftance, or fome other line 

 equally inclined ; and the angle meafured has been, from the 

 point where the inverted image is terminated by the water, to 

 that part of the oar itfelf which appears to be diredly above it. 

 (The apparent magnitude of ec. Fig. 1. PI. 3.) 



The eight firft angles were taken with a mother-of-pearl 

 micrometer in the principal focus of my telefcope, and are not 

 fo much to be depended upon for accuracy as the fucceeding 

 eight. Tfaefe laft were meafured with a divided eye-glafs 

 micrometer, and confequently are not liable to any error from 

 unfleadinefs of the inftrument or objefl. 

 It is not likely From the foregoing obfervations we learn, that the quantity 

 ■variation of the o^ refradion over the furface of water may be very confiderable, 

 dip happens out where the land is near enough to influence the temperature of 

 the air. At fea, however, fo great differences of temperature 

 cannot be expected ; and the increafe of dip caufed by this 

 variation of horizontal refraction, it is to be prefumed, is not fo 

 great as in the confined courfe of a river ; but, if we contider 

 that it may alfo be fubje6l to an equal diminution from an op- 

 pofite caufe, and that the horizon may even become apparently 

 elevated, there can be no queftion that the error in nautical 

 obfervations, arifing from a fuppofition that it is invariably 

 according to the height of the obferver, ftands in need of 

 correction. 

 Mr. Huddart's The remedy employed by Mr. Huddart,* of taking two 

 leme y. angles of the fun from oppofite points of the horizon at the 



fame time, and confidering the excefs of their fum above 180° 

 as double the dip, muft without doubt be efFedual ; but, from 

 caufes which he affigns, it is practicable only within certain 

 limits of zenith diftance ; for, where the zenith diftance is 

 fmall, and the changes of azimuth rapid, there is required con- 

 liderable dexterity and fteadinefs of a tingle obferver who at- 

 tempts to turn in due time, from one obfervation to another ; 

 and, when it exceeds 30°. the greater angle cannot be mea- 

 fured with a fextant, and confequently his method is, with 

 that inftrument, of ufe only in low latitudes. 



* Phil. Tranf. for 1797, p. 40. 



On 



