1880.] Blot and others on horizontal refraction. 23 



I may as well insert here one observation of my own, of a con- 

 tact of the Sun's lower limb with the sea-horizon, which I timed 

 accurately, at Calais, on the sands west of the pier, with my eye at 

 an elevation of about 18 feet above the level of the water, on the 

 evening of Sept. 18, 1877, at about 5h. 56m, 40s. G.M. T. The sea 

 being calm, and the horizon and sky perfectly clear, I consider it 

 worth appending to the observations of which I have given an 

 abstract. 



Assuming Ivory's theory of refraction, which can be employed 

 for greater zenith-distances than Bessel's, to be approximately 

 correct, I find the zenith distance of the Sun's centre, assuming 

 the usual allowance of 4' 20" for dip, to have been when its lower 

 limb touched the horizon 90° 23' 27" : whereas, calculating its 

 place by theory, it ought to have been 90° 24' 27", or 1' greater. 

 This would seem to shew that at the moment when I observed the 

 Sun, the depression of the visible sea-horizon, was that much 

 greater than the normal, or 5' 20" about. Now at Dunkirk, in 

 the first story of the house, where the observer's elevation was 

 much the same, viz. 28 — 35 feet, when the sea was warmer 

 than the air the dip was greater than the normal, (1) 1'. 49", 

 (2) 1'30", (3) 1'38", (4) 37"; but when the sea was colder than 

 the air the dip, from the same position, was less, (1) 1' 20", 

 (2) 1' 25", (3) 1' 37", (4) 1' 25". When I saw the Sun setting^ 

 the evening chill was rapidly coming on, and so the temperature of 

 the air probably falling faster than, if not below, that of the sea : 

 the result thus agreeing in a certain degree with that observed at 

 Dunkirk. I had no thermometer to note the temperature of the 

 sea water, but that of the air, as well as the height of the baro- 

 meter, I got on the pier. 



It results clearly from these observations — I mean of course 

 mainly those of Biot, that the temperature of the atmosphere can 

 be a very important factor in depressing or elevating the visible 

 sea-horizon, and probably also the observed altitude of any distant 

 object, with reference to the true horizon at the point of observa- 

 tion. In observing from the higher points, allowance was made 

 for the curvature of the earth, as is stated by Biot at p. 151 of 

 his work. 



I will now proceed to consider the effect of refraction of the 

 position of the Sun on or near the sea-horizon, as observed by 

 Le Gentil at Pondicherry in 1769. His observations will be found 

 given in considerable detail in the Memoir es de VAcademie des 

 Sciences, 1774, pp. 330 — 350. 



Le Gentil prefaces his remarks by saying that at the beginning of 

 the 18th century there was an impression that refractions at the Pole 

 were much greater (perhaps double) what they are at the Equator, 



