720 The Account of a 



it variable, are as likely to predominate when the angles of de- 

 pression or elevation are observed from low stations as when 

 observed from high ones, we may be enabled to make some 

 general deductions.* 



When the instrument formerly made use of by General Roy 

 was intrusted to my care, I possessed the means of deter- 

 mining, in a more accurate manner than had yet been done, 

 the refractive power of the air near the horizon. To devote 

 much time to it, has not, as yet, been in my power ; because a 

 more rapid extension of the survey was an object of greater 



* As many instances of strong atmospherical refraction have been related, and inge- 

 niously accounted for, in some of the late publications of the Royal Society, I think it 

 right to mention, by way of note, a very extraordinary instance of its variability. 



In the month of June, 1795, wnen tne instrument and party were stationed at 

 Pilsden Hill, in Dorsetshire, on a particular day, at about the hour of four, I em 

 ployed myself in observing the angles of depression or elevation of the surrounding 

 hills. After I had done all that was necessary in this matter, I turned the telescope to 

 Glastonbury Tor, and observed the depression of it. The air was so unusually clear, 

 that, desirous of proving to a gentleman then with me in the observatory tent, the 

 excellence of the telescope, I desired him to apply his eye to it : this he did, and, agree- 

 ably to a desire he expressed, I again took the depression of the upper part of the old 

 building, which I was enabled to do with great accuracy, and found it 2" different ; 

 the first being 3o',o", and the last 3Q',2". The unusual distinctness of this object, led 

 me to keep my eye a long time at the telescope ; and, whilst my attention was engaged, 

 I perceived the top of the building gradually rise above the micrometer wire, and so 

 continue to do, till it was elevated io',45" above its first apparent situation ; it then 

 remained stationary, and as night drew on, the object became indistinct. The follow- 

 ing evening, I observed the depression again, and found it 2c/,5o". To what cause this 

 extraordinary change in the refraction could be owing, I am at a loss to conjecture. 

 The former part of the day had been warm, with little wind, and cloudy. The thermo- 

 meter, at the time of observation, was 65 , and continued stationary for a considerable 

 time. The sky was cloudy, but yet, as I have before observed, the air was remarkably 

 clear. The top of Glastonbury Tor, I suppose, is about 200 feet from the surface of 

 Sedgemoor, over a considerable tract of which, the line joining Pilsden with that object 

 passes. The gentleman of whom I speak, as being with me in the tent, was Captain 

 Darcy, of the Royal Engineers, who, no doubt, well remembers the circumstance. 



