726 The Account of a 



On examining the refractions obtained on the first arc, we 

 perceive them to have been tolerably regular from 3 o'clock 

 till 8 ; the mean being -J _ part of the contained arc. The 

 height of Whiteham Hill is 576 feet, and that of White Horse 

 Hill 893 feet, above the level of the sea : the ray passes, there- 

 fore, through a tract of air considerably elevated, as the country 

 between the stations is, for the most part, flat and low. 



The air is not often clear enough, or sufficiently free from tre- 

 mulous motions, for these delicate observations. On the present 

 occasion, however, the state of it was highly fit for the purpose; 

 and, as care was taken, I am of opinion an error of more than 

 3", taking that of the arch of altitude into the account, cannot 

 have obtained in any of the angles. The refractions at 9 

 and 1 o o'clock are less than at the preceding hours ; but 

 this does not appear to have been owing to any change in the 

 refractive power of the air throughout the whole extent of the 

 ray, because the depression of Whiteham Hill, from the other 

 station, varied little at those hours. These changes in the ob- 

 served angles of elevation at Whiteham, (44" and 42" being the 

 differences, ) without corresponding ones at White Horse Hill, 

 prove that some partial alteration, from floating strata, had taken 

 place in the refraction near the former station. Whoever con- 

 siders the matter, must perceive a case may be constructed in 

 which this will take place, causing a great variation in one of 

 the angles, whilst the other apparently remains the same : and 

 this suggested the idea, that to afford any accurate conclusions 

 in this way, a long series of observations would be necessary. 

 It furthermore appears, that dew could not have caused these 

 differences at Whiteham Hill, since the same cause would 

 equally operate to vary the observed angles at W T hite Horse 

 Hill ; but those remained nearly the same. 



