as applied to the measurement of Heights, 337 



periments. Its level below station A I ascertained with the 

 nicest accuracy, by levelling with a theodolite by Troughton, 

 to be 141.1 feet. From a mean of some single observations 

 made with the sympiesometer in the spring of the same year, 

 I obtained nearly a similar result. The observations now to 

 be detailed were made in the height of summer, and their great 

 discrepancy in some cases from the real height indicate, I fear, 

 some more fundamental want of compensation than in the 

 mere horary variations of temperature. Without regarding 

 the order of dates, we shall commence with the most irregular 

 sets of observations, and go on to those more accordant with 

 the result by levelling, and more accurate in themselves. The 

 mode employed was, after making the observation at the 

 higher station, to proceed as quickly as possible to the other, 

 and after the cork which closes the cistern of oil had been open 

 generally from two to four minutes, the register commences, and 

 was continued every five minutes for an hour, with the purpose 

 of discovering the ultimate time required for the self-adjust- 

 ment of the instrument. We shall arrange them in tables ac- 

 cording to each day's observations, and make our remarks on 

 each. The height of the sympiesometer is given in inches, and 

 also in fathoms, as marked on the scale, and the column of 

 DilF. indicates the difference of level in fathoms indicated be- 

 tween the two stations, but which of course requires the usual 

 thermometric correction for the temperature of the air, as in 

 measurements by the barometer. 



VOL. X. NO. II. APRIL 1829. 



