ATMOSPHERE OF GREAT BRITAIN AND ON THE CONTINENT. 657 



1892 show the air to be clear when the amount of dust is small, but they offer no 

 information as to the effect of a large number of particles. This point is better brought 

 out in the 1893 observations, to which we shall now refer. 



The first day in the 1893 observations when the mountain was free of cloud was the 

 26th of June ; the transparency was 4, and Diagram III. shows that the amount of dust 

 at both levels was small. On the 30th the transparency was 3. " Fog and haze were 

 observed in the valley " in the morning. It will be seen from the diagram that the 

 amount of dust was small at both stations in the morning, but rapidly rose at low level. 

 The Ben Nevis report for 7 p.m. says, " Haze bank to S.E. and S." This haze appears to 

 have begun to affect the dust readings shortly after mid-day ; the number of particles was 

 fairly high at 7 p.m. and was very high at 10 p.m., at which hour the transparency was 

 reduced to 1, there being " a thick haze all round." On the following day, the 1st July, 

 observations were fortunately again possible, this being one of the days in the 1893 dust}? - 

 periods. On this occasion, though the air was dry, the transparency was never more than 

 1. The diagram shows that at both stations the amount of dust was great on this day. 

 On the 2nd the amount of dust was less and the transparency had increased to 2. 

 On the 3rd the amount of dust was still much the same as on the 2nd, and the 

 transparency was again 2. The 5th had a transparency of from 2 to 3, and the amount 

 of dust at both levels was not great. In the Ben Nevis report for this day, " Haze in 

 valleys " occurs two or three times. This hazy impurity does not seem to have penetrated 

 to the Ben or to Kingairloch on this day. Next morning, that is on the 6th, the trans- 

 parency was still 3, and remained at that figure till mid-clay, and during that time the 

 amount of dust was low at both stations. In the afternoon, however, the transparency 

 fell to 1, and the number of particles had increased to a high figure at both levels by 

 that time. The low transparency and great amount of dust observed this afternoon were 

 probably due to the arrival that day of the haze observed the day previous. On the 7th 

 the transparency was only 1, and the amount of dust was very great at both stations. It 

 became clearer on the morning of the 8th, being then from 1 to 2, and the amount of 

 dust was also less, but clouds came in the afternoon and closed out the view. On the 

 11th the transparency was 2, and dust at both levels above the mean. The 15th was a 

 clear day, transparency being 3 to 4, and the diagram shows that the amount of dust 

 was small, if we omit the doubtful high readings on the Ben, also the abnormal ones in 

 the afternoon at low level. It was clear in the early morning of the 16th, when the 

 amount of dust was small, but no transparency observations were possible after 7 a.m. 

 The early morning of the 18th was also very clear and the amount of dust very low. 

 After that date mist covered the mountain top till the close of the observations in the 

 diagram for 1893. 



The conclusions arrived at from the 1893 observations are, that on all the days on 

 which the air attained its maximum clearness the number of particles was low, and on all 

 the days on which it had a minimum transparency, or maximum of haze, the number of 

 particles was great, the least transparency being observed from the 30th June to the 1st 



