^56 Altitudes ofMountainSi ^c. visible Jirom 



the utmost skill of the observer to produce consistent results : 

 yet I am at a loss to conceive how I should have estimated 

 the error at 25", unless, as in the case of the proof level, I had 

 erroneously noted the double of the quantity. By referring to 

 the preceding list of refractions it will be evident, that as the 

 refractions for the smaller arcs are in general considerably in 

 excess, the constant error of the instrument must have been 

 sensibly over-rated. 



The subjoined list of refractions, determined (with few ex- 

 ceptions) in the course of last year, will require some little ex- 

 planation. 



The stations are arranged in the order of their altitude 

 (which is affixed), and are followed by the date, time, and num- 

 ber of observations, together with sundry meteorological re- 

 marks. 



The first column contains the initials of the stations ob- 

 served: — the second, the partiadar oh]&ct of pointure: — the 

 thii'd, the arc: — the. fourth, the number of observations : — the 

 ^^th, the refraction resulting from the difference of the true 

 and observed angle : — the sixth, the bearings of the stations : — ■ 

 the seventh, the differeiice of altitude as deduced by making 

 use of one half of the sum of the observed refractions, or of 

 1-1 8th when the observations were not reciprocal. The last 

 column gives the difference corrected in the manner described 

 in my last. 



An asterisk (*) prefixed to a I'efraction denotes that it is sup- 

 posed to exceed the mean J (l-18th) in consequence of the 

 mountain observed being flat or slightly rounded. 



^^'^hen the ground was not the object of pointure, the re- 

 fractions appear to have been in most cases remarkably small. 

 They are distinguished by a prefixed f . 



In the direction of the excess of matter the zenith distance 

 wiil be measured too great, and the refractions will fall short 

 of its mean value. By using a mean quantity in calculation 

 tlie determined differences of altitude will be in defect, or in 

 excess, as the objects observed are higher or lower than the 

 station itself. In cases that admit of little or no doubt, the 

 mark + denotes that tha plumb-line will have been deflected 

 towards the object observed. The mark — indicates a defect 

 of matter in the direction of the station to which it is prefixed. 

 It is supei-fluous to insist upon the extreme difficulty of deter- 

 mining the apparent zenith of the observer when the instru- 

 ment is placed on the brink of a deep, broad and nearly pre- 



X The mean refractioTi was foiind by trjing what proportion half the 

 diftcrcncc of tlic sum of all the observed depressions, and the sum of tlie 

 observed clcvatioui + the contained arts, bore to the sum of the said arcs. 



cipitous 



