MR AITKEN ON ATMOSPHERIC DUST. 483 



if the morning was sunny, but the increase coming later if the morning was cloudy and 

 followed by sunshine, and not appearing at all if the day remained clouded. 



The effect of the dust on the transparency of the air was also noted during these 

 observations made at Ben Nevis and Kingairloch, and the general result arrived at is 

 that on all clear days the number of particles was small, and on all thickly hazed days 

 the number was great. Take, for instance, the 16th of July 1892. During the whole 

 of that day the numbers were very low at both stations. At high level the number 

 was 160 per c.c, and at low level 200 per c.c. In the Ben Nevis Observations it is 

 noted : " Ireland seen, only a thin haze." As Ireland is 125 miles distant, the air must 

 have been very clear to give only a thin haze at that distance. The observations made 

 at Kingairloch on that day show the air had its maximum clearness also at low level. 

 Many similar observations might be quoted to show that the air is clear when the number 

 of particles is few, and also to show that it is thick when the number is great, all of which 

 point to the conclusion that dust is the principal cause of the haze in the atmosphere. 



The hazing effect of dust is confirmed by the observations on the Rigi Kulm. The 

 air there was never clear unless the number of particles was small. The amount of haze 

 at that station was estimated by the clearness of Hochgerrach, a mountain seventy miles 

 distant from the Rigi. During five visits to the Rigi Kulm, Hochgerrach was visible 

 thirteen times. On eight of these occasions the mountain was only one-half to one-fifth 

 hazed, and the number of particles was from 326 to 850 per c.c. It was seen five times, 

 when the number varied from 950 to 2000 per c.c. ; but on these occasions the mountain 

 was little more than visible, and it was never seen when the number of particles was a 

 little over 2000 per c.c. 



The hazing effect of dust has been worked out in detail by me from observations 

 made at Kingairloch, Rigi Kulm, and Alford in Aberdeenshire.* In working out the 

 relation between the transparency of the air and the different amounts of dust in it, it 

 was found necessary first to arrange the different observations according to the humidity 

 of the air at the time of observation, as it was found that the humidity had a marked 

 influence on the transparency. It is not supposed that humidity, as humidity, has any 

 effect on the transparency. Water vapour, as such, seems to have no hazing effect, 

 because we know we can have very transparent air with a high humidity if the number 

 of dust particles be small. The effect of the humidity would seem to be to increase the 

 size of the particles, so increasing their hazing effect. Some water seems to be con- 

 densed on the particles even in dry air, but less is condensed the dryer the air. The 

 separate observations were accordingly arranged in different tables. In one table were put 

 all the observations taken when the wet bulb depression was from 2° to 4° ; in another, 

 the observations when the depression was from 4° to 7° ; and in a third, all observations 

 when the depression was over 7°. The observations were arranged in the different 

 tables in order, beginning at the top of the column with the observation having the 

 smallest number of particles, and in another column alongside entering the limit of 



* See Proc. Roy. Soc. Edin., vol. xvii., 1889-1890, p. 193. 



