REPORT ON ATMOSPHERIC CIRCULATION. 41 



On comparing this latter table with the empirical one, it was seen that the two 

 agreed throughout in giving the same differences between two different sea level 

 pressures at the same air temperatures. But the two tables differed essentially when 

 compared as to their differences for the same sea level pressures at different air tem- 

 peratures. At the air temperature of 45° the two tables agreed, at lower temperatures 

 the corrections from Laplace's formula were too large, and at temperatures higher than 

 45° too small. It was found that, when the additions to the corrections in the Laplace 

 table for air temperatures lower than 45° were reduced by one-sixth, and the subtractions 

 from the corrections as the temperature rose above 45 were also reduced by one-sixth, 

 the two tables were virtually identical. It may be noted here that the differences 

 among the corrections for height arising from the varying air temperatures thus deduced 

 from the Ben Nevis Observations substantially agree with the differences in Hazen's 

 Table for the reduction of Air Pressure. 1 



A table was then constructed from Laplace's formula for a sea level pressure of 

 30'000 inches for latitude 45°, and for air temperatures from —20° to 90°, and for 

 heights up to 8000 feet. To the figures of this table were applied corrections for the 

 different air temperatures, in accordance with the results of the Ben Nevis Observations. 

 The result is given in Table V., which has been used in reducing the barometric means 

 of Table VI. to sea level. The table is, however, only regarded as a provisional one, 

 giving tolerably good approximations to the true corrections for height. 



But the really serious difficulties encountered in reducing barometric observations 

 to sea level are presented by the air temperatures, and unless these difficulties are kept 

 steadily in view, no little confusion will be the result in representing the course of the 

 isobars. The more serious of these difficulties are experienced in dealing with stations 

 situated in deep, narrow valleys, and stations on elevated plateaux. 



This is well shown by the observations made at Obirgipfel in the Tyrol, which is 



a high level station on a peak 6706 feet high, and at Klagenfurt, about 7 miles 



distant, in a deep valley adjoining, at a height of 1437 feet, there being thus a 



difference of 5269 feet in height between them. Now the differences of temperature 



between the monthly means of these two situations for the five years 1880 to 1884 are 



these, the figures showing the excess of the temperature of Klagenfurt above that of 



Obirgipfel : — 



0-7, 5-8, 12-6; 19-4, 22-0, 21-2; 18-5, 176, 1G4; 139, 7-7, 61, 



and for the year 13 "4. Now, since the station at Obirgipfel is situated on a true 

 peak, it follows that the temperature there recorded will closely approximate to the 

 temperature of the free atmosphere at that height. But at Klagenfurt it is far other- 

 wise, for being situated in a deep narrow valley, the night and winter temperatures, as 

 already explained, are greatly too low, and the day and summer temperatures are too 

 high. The mean winter temperature at Klagenfurt is only 4° '2 lower than that of the 



» "Washington, 1882. 



(PHYS. CHEM. CHALL. ESP. PART V. 1889.) " 



