178 REPORT—1890. 
pressures at the two stations is virtually a temperature effect. As the 
temperature falls to the annual minimum in winter, the air contracts, 
and a portion of it consequently falls below the level of the barometer at 
the top, thus reducing the readings there, and increasing the differences 
between the two barometers. The difference then reaches 4°632 inches, 
the maximum for the year. On the other hand, as temperature rises, a 
portion of the atmosphere is raised above the level of the higher baro- 
meter, thus increasing the pressure there, and lessening the difference to 
4485 inches in July, the minimum of the year. The difference between 
the maximum and minimum is thus 0°147 inch. For these months the 
mean temperatures of the stratum of air between the top and bottom of 
the mountain are respectively 30°°5 and 48°°5. Hence the vertical dis- 
placement of the mass of the atmosphere for a temperature difference of 
18°:0 is represented by a barometric difference of 0:147 inch. The sea- 
level pressures in these months are, however, respectively 29°876 inches 
in February, and 29°841 inches in July. If, then, we assume the sea- 
level pressure of July to be the same as that of February, viz. 29-876 
inches, the difference between the top and bottom pressures would be not 
4485, but 4°490 inches. From this it follows that the vertical displace- 
ment for a temperature difference of 18°:0, and at the same sea-level 
pressure, is 0142 inch. 
Sosgesetsssssaiasssssazesis 
EEE HEH He 
Srdiecnetiietirentisedi 
ri ty cae, PtH oni aauaae 
amass sas 
Annual curve of the differences of barometric readings for high and low level stations 
(Ben Nevis and Fort William). 
In order to determine the curve of the table of the barometric differ- 
ences, it is convenient that these should be reckoned from the mean point, 
or say 4560. When so treated, the differences are :— 
May —-016 November + *024 
June —*050 December + °043 
July —:075 January + ‘070 
August —:073 February + ‘072 
September — ‘057 March + 067 
October —-019 April + *031 
These quantities being laid down as vertical ordinates, with Time as 
the horizontal ordinate, it was evident to the eye that the curve was a 
projection of the curve of sines. The difference between the extreme 
and mean values in 455 parts of an inch is 75. Hence, if a be the time 
expressed in arc, and 68 the differences in the preceding table, we have 
als ss= sin a. 
