Nen' Tables for Finding Heights by the Barometer. 73 



or on a separate ring which can be set to zero at starting, 

 and thus save the trouble of taking the differences. This 

 latter plan, although more convenient, is not so accurate as 

 the former ; but in every case, if the best result is desired, 

 the readings should be taken from the scale of inches or milli- 

 metres, as it is uncertain to what temperature the altitude 

 scale is adapted. I have measured some of the English ones, 

 and find the temperatures they represent to vary from 47° 

 to oT, but the French one mentioned before has its scale 

 adapted to 70° Fahr. The small tables supplied by instru- 

 ment makers, which have no temperature factor, are 

 generally adapted to about 50° ; while the larger tables, 

 which contain the factors for temperature, latitude, and 

 decrease of gravity, are usually given for 32°, the freezing 

 point. The adoption of so low a temperature as 32° is 

 very inconvenient for Australia, where the mean tempera- 

 ture is. much higher, as it necessitates a very large cor- 

 rection, and as this correction is frequently neglected, the 

 results must be very inaccurate. I have therefore thought 

 it desirable to compute a new table, in which the mean 

 temperature and middle latitude of this part of Australia 

 should be used. Calling B the height of the barometer^ 

 and t the temperature of the air at the lower station, 

 B' and t' the same quantities for the upper station, L the 

 latitude, and A the difference of height in feet between 

 the two stations. Laplace's formula, leaving out the 

 factors depending on the decrease of gravity, may be 

 written : 



A = 60158-6 (log. B- log. B') x A + t + t' - 64\ q.qq265 cos 2 L. 



Taking now T and T' equal to 60°, and L = 37', and 

 substituting them in the above equation, we get 



A = 63948-6 (log. B- log. B'). 



The mean height of the barometer at the Melbourne 

 observatory, 91 feet above the level of the sea, reduced 

 to 32° Fahr., is 29 "931 in. ; increasing this by 0-07 to bring 

 it up to 60° Fahr., nearly the mean temperature of Mel- 

 bourne, and by O'lO to reduce it to sea level, we get 

 oO'JO ; with this value in the above equation we should 

 find 63948-6 log. B = 94552'26. It would be more con- 

 venient, however, to have it represented by o, because 

 the tabulated values corresponding to the height of the 

 Imrometer will at once show the height above the le\'el 



