80 Proceedings of the Royal Society of Victoria. 



determinations by barometer were all too low, the best 

 being four feet out, the worst eight feet, and the average 

 six feet in error — Laplace's table being used in the 

 calculation. 



A second experiment at the same time, consisted in 

 comparing three points within half a mile of the Pipehead 

 dam, and within 1 00 feet of its level with the dam itself. 

 The level ot two of these points was given within two feet 

 of the truth by a single observation, while the third which 

 was not so well observed, owing to night coming on, was 

 five feet in error. 



A third experiment which has been frequently performed 

 with the surveying class at the Universit}^ consists in 

 taking the height of the gallery on the tower of Ormond 

 College. This is 130 feet from the ground, and has been 

 many times measured by instruments 1 and 2. The result 

 has always been slightly below the truth, usually about 

 125 feet. I do not remember it ever having been below 

 120 or above 130. 



On the 1st inst., having Mr. White's lemarks in view, 

 and having for other purposes to pay a visit to a place 

 called Riverview^, near the Yarra Glen Railway Station, I 

 determined to make a careful test of the three barometers, 

 and compare the result when fully worked out and all 

 proper corrections applied, with the known difference of 

 level by actual levelling. I therefore set the barogra]:)h 

 going at the University, and carefully reading the three 

 instruments at the Princes Bridge Railway Station, took 

 them to River view, and read them as other engagements 

 permitted, eight times in forty-eight hours. Thus I obtained 

 twenty-four readings, which, when compared with the Princes 

 Bridge readings, corrected by reference to the barograph and 

 worked out by Laplace's table with instrumental and tem- 

 perature corrections, gave the following results : — 



Mean 235-7 232-4 229-6 



