ATMOSPHERE OF GREAT BRITAIN AND ON THE CONTINENT. 641 



these observations are those made on the 9th July 1891 and on the 14th and 15th 

 July 1893. 



On the 9th July 1891 the morning reading was 295 per c.c., at mid-day it was 3700, 

 and at 3.30 p.m. it was as high as 7600. The 1893 observations are still more extra- 

 ordinary. On the 14th the morning reading was about 90 per c.c., at 1 p.m. it was up 

 to 12,250, and at 2 p.m. 13,500 per c.c. After 1 p.m., five observations were made, which 

 showed the number to diminish regularly to 2100 at 7 p.m. and to 378 at 9.30 p.m. 

 Next morning (the 1 5th) the number at 9 a.m. was 420 ; at 1 p.m. it was very irregular 

 and very high, and went on rising till 4 p.m., when it was as high as 9325, after which it 

 steadily got less and less, and at 9.30 p.m. the figure was 1412 per c.c. 



It will be noticed in all the abnormal cases, wheu the number went high during the 

 day, the number was very low in the morning, generally two or three hundred, and in one 

 case less than one hundred ; and also after attaining a maximum number in the after- 

 noon, the number always fell fairly regularly to a small figure, but not generally to so 

 small a number as in the morning. A great many observations were made on the two 

 days of most abnormal readings, namely, the 14th and 15th July 1893, and, as will be 

 seen from Diagram III., the rise and fall are quick and fairly regular, the line drawn 

 through the different observations for each day making a fairly regular curve. 



It will be further noticed from the three diagrams that on many days with N.W. 

 winds there was no tendency for the numbers to rise in the afternoon ; for instance, 

 the numbers remained fairly steady with N.W. winds on the following days: in 1891, 

 on the 7th, 8th, 10th, 20th, 22nd, 24th, 25th, 26th, 28th and 29th July and 1st 

 August ; in 1892 the numbers were low or steady all day on the 1st, 4th, 5th, 7th, 

 8th, 19th, 22nd ; and in 1893 they were low all day on the 22nd, 23rd, 24th, 29th June 

 and 20th July. The previous observations of abnormally high readings in N.W. winds 

 thus receives very ample confirmation by the observations made during the last three 

 years. From these last observations we see that while on three out of every five days 

 with N.W. winds the numbers remained low all day, yet in two out of every five the 

 numbers went high at some time of the day. 



A marked peculiarity of all the days of abnormally high readings is that on all of 

 them the air was "clear" to "very clear" as will be seen from Tables I., II. and III. 

 With so large a number of particles as was frequently observed on the abnormal days 

 there is always a great amount of haze under ordinary conditions. The only peculiarity 

 noticed in the haze was that the whitish appearance, w r hich generally accompanies large 

 numbers, was absent, and the haze had a peculiarly fine blue tint seldom seen. 



Though the existence of these abnormal readings is well established by the last 

 three years' observations, yet there is great difficulty in offering an explanation of them. 

 In looking for their cause, the first thing to be done was to see if there was any 

 relation between the weather conditions and the exceptional readings. Did they come 

 with all kinds of weather, or were they associated with any particular conditions \ For 

 the purpose of working out this point I have been in the habit of recording the condition 



