EXPLANATION OF THE PLATES-, X)eX>iu 



io afcribe thefe optical efFeds to the fimple laws of the re- 

 fradion of the Atmofphere. According to Bougucr, a 

 .well-known Academician, in his Treatife on Aavigatiofjf 

 book iv. chap. 3. fedion 3. *' Refra£lion elevates the ftars 

 ^' in appearance ; and we are af Fared, by an infinite num- 

 *' ber of certain obfervations, that when they appear to us 

 ** in the Horizon, they are, in reality, 33 or 34 minutes 



*' under it In regions where the air is more denfe, ths 



*' refradions mud be fomewhat ftronger, and they are, 

 ** like wife, every thing fclfe being equal, fomewhat greater 

 ** in Winter than in Summer. In the pra6lice of naviga- 

 ■" tion that difference may be entirely negleded, and per- 

 ** petual recurrence may be had to the fmall table placed 

 *' on the margin." 



You fee, in facl:, at this part of his work, a fmall table, 

 in which he lays down the greatefl refraélion of the Sun in 

 ihe Horizon, at 34 minutes, for all the climates of the 

 Globe. But how came it to pafs that Barents fhould have 

 feen the Sun above the Horizon of Nova Zembla, on the 

 24th of January, in the fign of Aquarius, at five degrees, 

 twenty-five minutes, whereas he ought to have been there, 

 in fixtecn degrees, twenty-feven minutes, in order tj be 

 perceived in the fevent3^-fixth degree of northern Latitude, 

 where Barents then was ? The refradion of the Sun, then, 

 above the Horizon, was nearly two degrees and a half, 

 that is, four times as great, nay, more than Bouguer fup- 

 pofes it to be, as he afligns only thirty-four minutes, or 

 nearly, for every climate In general. 



Barents^ in truth, was very much aflonifhed to fee the 

 Sun fifteen days fooner than he expedled \ and he could not 

 be perfuaded that it actually was only the 24th of January, 



VOL. I. c but 



