96 



These heights are calculated from the levels of Richmond-place, and 196 

 feet 1 inch must be added to give the height of either of tliese hills above the 

 level of the sea. With respect to Backbury, I wish to mention that the fir trees 

 do not stand upon the highest point of the hill. I hope to measure this hUl 

 again, and to obtain permission to place the barometer on the most elevated 

 part of the camp. 



The altitiides given above are, I am satisfied, very close approximations 

 to absolute truth ; for the measurements were made with care for all the hills, 

 and repeated with satisfactory results for Acornbm-y, Credenhill, Hough Wood, 

 and Dinedor. Garway and Backbury have each been measxired once only ; but 

 on Garway several observations were made at short intervals, answering to a 

 corresponding number made at Hereford, and the calculations were worked 

 out separately by the Rev. R. Dixon, Mr. Easton, and Mr. Cooke. 



Measurement of altitude by standard barometers is wonderfully accurate 

 provided certain conditions are observed. In the first place, of course, it is 

 necessary that the barometers should be of the best description. Their errors 

 must be known. They must read to the 500th of an inch. In the second place, 

 it is important to possess Uco instruments, so that the pressure of the atmosphere 

 at the upper and lower stations may be determined at the same moment. In 

 the third place, it is important to determine the air temperatvire by means of 

 good thermometers so placed at both stations as to furnish correct information 

 on this point. In the fourth jilace, it is desirable to reduce the horizontal 

 tlistance between the upper and lower stations as much as possible. 



Mr. Drew, in his "Practical Meteorology," informs us that on one occasion 

 the following experiment was made at Greenwich :— Observations were first matte 

 in the transit-room of the Royal Observatory, and at the base of the statue of 

 George II., in Greenwich Hospital, to determine the difference of altitude, and 

 then the barometer work was tested by means of the si^irit-level. The results 

 were as follows: — By barometer 135 '4 feet; by levelling, 135 '57 feet. 



The measurement of Etna affords a striking proof of the value of the 

 barometer as an instrument for ascertaining the heights of mountains. This 

 volcano was measured by Captain Smyth, in 1815, by trigonometrical survey, 

 and found to be 10,874 feet high. In 1824, Sir John Herschel, ignorant of 

 the height assigned to the mountain by Captain Smyth, measured its altitude 

 by barometer, and found it to be 10,872 feet C inches ! 



I must, however, allude to one source of error which may in some small 

 degree vitiate our measurements, and it is this : At the time the measurements 

 were made the Woolhoi^e barometer had not been tested, and in correcting 

 for error of instrument we depended partially on the correction given by the 

 maker and partially on a correction derived from a comparison of this 

 barometer ^vith my own : but any error arising from tliis source must be very 

 small indeed. 



1 



