24 Br J. B. Pearson, On the experiments made by [Nov. 22, 



and mentions that Maupertuis had dissented from this view, adding 

 that the well-known asserted sight of the Sun by the Dutch 

 in Nova Zembla, on Jan. 24, 1597, cannot be relied on as an 

 element to be taken into account in the discussion. 



He then states that when at Port Louis (20° 10' S. Lat.), in the 

 Isle de France or Mauritius, he noticed in winter that the Sun 

 never set behind the visible sea-horizon, but a false horizon, four 

 to five minutes (arc) higher. This observation he could not verify 

 in summer, the sea-horizon being then interrupted by land ; and at 

 Manilla, where Ee went to observe the transit of Venus in 1761, a 

 suitable place for similar observations could not be obtained, but 

 when he had established himself at Pondicherry, on the Coromandel 

 coast, he soon found that the peculiarity which he had noticed 

 at Port Louis entirely repeated itself: the Sun invariably rising in 

 winter, not from behind the sea-horizon, but from as it were a 

 bank of mist about 5' higher, and occupying only about 35 seconds 

 of time in reaching the horizontal wire of his telescope when 

 adjusted at zero : whereas in summer, it rose always from behind 

 the natural sea-horizon, and occupied nearly a minute, 59 seconds 

 on one occasion, in reaching the horizontal wire of the telescope, 

 and with this difference in brightness, that the darkened glass was 

 required in summer as soon as the Sun's upper limb had passed 

 the horizon, while in winter it could be viewed with the naked 

 eye up to one degree of elevation. He ascribes this phenomena to 

 the condensation of the atmosjmere, even within the tropics during 

 winter. 



He then gives his actual observations, very fully recorded, of 

 which I have analyzed and computed such as serve my purpose, as 

 follows : 



He commences by ascertaining the depression of the sea- 

 horizon, which on Jan. 7, 1769, he found to be 10' 50" at an 

 altitude of about 40 feet, or, allowing for an error of 1' 58" in 

 the graduation of his instrument, a quarter circle with telescojDe 

 apparently similar to that used by Biot, about 8' 52". As the 

 normal value would be about 6' 36", it seems that the horizon 

 appeared depressed about 2' 16" more than would be expected, and 

 he says that he had the same result from an observation on the 

 following day; and apparently also, during the summer, though he 

 gives no more actual figures. 



He then gives the actual moments of the Sun reaching the 

 wire of his telescope, of course at a Zenith distance of 89° 58' 2", for 

 a number of days during January and February, and also June 

 and July. I have computed these, with the aid of the Nautical 

 Almanac for 1769, employing Ivory's tables for refraction, and 

 assuming the temperature at sunrise to have been 60° F. in winter, 



