VOL. LXXX1X.] PHILOSOPHICAL TRANSACTIONS. 437 



But, besides the usual refraction which affects the rays of light, the atmosphere 

 over the sea is sometimes found to be in a state which refracts the rays in such a 

 manner as to produce other images of the object, which we will call an effect from 

 an unusual refraction. In the Phil. Trans, for 1797, Mr. Huddart has described 

 some effects of this kind, which he has accounted for by supposing that, from the 

 evaporation of the water, the refractive power of the air is not greatest at the sur- 

 face of the sea, but at some distance above it; and this will solve, in a very satis- 

 factory manner, all the phenomena which he has observed. But effects very dif- 

 ferent from those which have been described by Mr. Huddart are sometimes found 

 to take place. These I had an opportunity of observing at Ramsgate last summer, 

 on August the first, from about half an hour after 4 o'clock in the afternoon till 

 between 7 and 8. The day had been extremely hot, and the evening was very 

 sultry; the sky was clear, with a few flying clouds. I shall describe the pheno- 

 mena as I observed them with a terrestrial telescope, which magnified between 30 

 and 40 times ; they were visible however to the naked eye. The height of the eye 

 above the surface of the water, at which most of the observations were made, was 

 about 25 feet; some of them however were made at about 80 feet from the surface; 

 and it did not appear that any of the phenomena were altered by varying the height 

 of the eye, the general effect remaining the same. 



The first unusual appearance observed, was that which is represented in pi. 8, 

 fig. ]. Directing my telescope at random, to examine any objects which might 

 happen to be in view, I saw the top of the masts of a ship a, above the horizon 

 xy, of the sea, as shown in the figure; at the same time also, I discovered in the 

 field of view, 2 complete images, b, c, of the ship in the air, vertical to the ship 

 itself, b being inverted, and c erect, having their hulls joined. The phenomenon 

 was so strange, that I requested a person present to look into the telescope, and 

 examine what was to be seen in it, who immediately described the 2 images, as 

 observed by myself; indeed they were so perfect, that it was impossible we could 

 differ in our description. On this I immediately took a drawing of the relative 

 magnitudes and distances of the ship and its images, which at that time were as 

 represented in the figure, as near as it was possible for the eye to judge; and it was 

 very easy to estimate them to a very considerable degree of accuracy. As the ship 

 was receding from the shore, less and less of its masts became visible; and, conti- 

 nuing my observations, in order to discover whether any, or what variations might 

 take place, I found that as the ship descended, the images b, c, ascended; but as 

 the ship did not sink below the horizon, I had not an opportunity of observing at 

 what time, and in what order, the images would have vanished, if the ship had so 

 disappeared. 



Being desirous of seeing whether the same effect was produced on the other ships 

 which were visible, I directed my telescope to another ship a, fig. 2, whose hull 

 was just in the horizon xy ; when I observed a complete inverted image b, the main- 

 mast of which just touched that of the ship itself. In this case, there was no 2d 



