BEN NEVIS AND FOBT-WILLIAM BAROMETERS. 515 



May; summer, June, Juh T , August; autumn, September, October, November; and 

 winter, December, January, February. 



The tables on pages 516 and 517 give the mean values at each hour of the day for the 

 four seasons of some of the relations of pressure and temperature. The yearly means 

 in all cases are the average of the four seasonal means. The mean values are first 

 given, and then in the next page the departures of each hour from the general mean of 

 the day, these latter values being as usual corrected for any difference between the 

 preceding and succeeding midnights. 



The first two items in the table are the differences of sea-level pressure and of 

 temperature between the two Observatories. It will be seen most easily in the 

 "Departure" tables that these are almost the complement of each other, the excess of 

 the one corresponding roughly to the defect of the other. But the correspondence is 

 not exact, and indicates that while the barometer differences are radiation effects 

 s}^mmetrical about noon and midnight, the temperature differences are symmetrical in 

 respect to the times of highest and lowest temperature. In autumn and winter the 

 time of the morning change of sign of the barometric differences is delayed owing to 

 the fact that the hills to the south-east of Fort- William prevent sunshine reaching the 

 Observatory there till some hours later than at Ben Nevis. Except for this the Ben 

 Nevis barometer is the lower from 7 to 18 o'clock, and the higher from 19 to 6 o'clock, 

 as appears in the Departure table, with one trifling exception at 18 in winter. More- 

 over, the curve of differences given by these figures is distinctly flattened at its maxi- 

 mum and minimum values, being in this respect unlike a diurnal curve of temperature, 

 and also unlike that of the differences of temperature given in the table. 



The third item in the tables is the mean hourly temperatures, the average of the 

 Ben Nevis and Fort- William temperatures. These are the temperatures that have been 

 used in reducing the Ben Nevis barometer to sea-level, and they have thereby been 

 assumed to be the mean hourly temperatures of the air between sea-level and the 

 height of Ben Nevis. If, however, instead of reducing the Ben Nevis barometer by 

 these temperatures we simply take the difference between the Ben Nevis barometer at 

 4407 feet and the Fort- William sea-level barometer, we can find from the reduction 

 table what the temperature at each hour ought to be to make the barometers agree at 

 sea-level. This is working the reduction table backwards ; instead of taking an 

 observed temperature and finding from it the amount to be added to the Ben Nevis 

 barometer to bring it to sea-level, we take the observed amount that the Ben Nevis 

 barometer differs from the sea-level one and find the corresponding temperature of the 

 intermediate air by that table. Carrying out this process for each hour of the day, we 

 get the temperatures given in the fourth section of the tables. These latter tempera- 

 tures may be regarded as the mean hourly temperatures of the air up to the level of 

 Ben Nevis on the days chosen for discussion, if we assume that the Ben Nevis baro- 

 meter, when brought to sea-level, ought to agree with the Fort-William one at each 

 hour of the day, and that the differences between them, when reduced in the ordinary 



