A GREEN SUN AND ASSOCIATED PHENOMENA. 395 



Meteorological. 



My attention was first called to the peculiar state of the atmosphere on 

 September 3rd, when making observations on atmospheric electricity. Observa- 

 tions made at 10 a.m. on August 31st showed a normal but rather low 

 electrification ; no observations were made on the 1st or 2nd September, the 

 electrometer being out of order. On the 3rd, after the instrument had been 

 carefully dried, observations were begun at 1.10 p.m., when the air potential 

 was found to be negative, the electrometer readings varying from —28 to —17 

 divisions. By 2.45 it had become positive, amounting, however, to only +6, 

 which was also the reading at 4.45. During the succeeding days numerous 

 readings were taken, and these show a remarkable state of electrification (see 

 Appendix A). The general results of the observations may be summed up 

 thus : — In the early morning the potential of the air was positive ; it continued 

 positive, but gradually decreased in strength till some time between 9 and 10 a.m., 

 when it fell to zero, and then became negative. The strength of the negative 

 charge varied greatly and with remarkable rapidity, and the maximum readings 

 were obtained either just before or during strong gusts of wind. The negative 

 charge continued till the sea breeze set in in the afternoon, when it became 

 positive, and continued so all night, and in all cases, I believe, the change of 

 sign coincided with a change in the direction of the wind. Very high readings 

 were at times obtained; in one case the reading was —459 divisions of the 

 scale ( — about 1900 Daniell cells or 2100 volts) at about 5 feet above the 

 ground, in bright sunshine with a clear sky, but with a gusty wind, which 

 raised clouds of dust. At other times there would be scarcely a trace of 

 electricity perceptible. In all cases in which the potential was negative it 

 decreased with height, but when it was positive it increased with height. 

 Another point worth noticing is that, when negative, the potential might vary 

 very much, but with two exceptions (and these are a little doubtful) it did not 

 change sign, as one finds so often in a thunderstorm. This state of affairs 

 continued from September 3rd to September 6th. On the 7th the potential 

 was positive all day, the day being cool and cloudy throughout. On the 

 8th it was negative for a short time ; on the 9th it was again positive all day ; 

 on the 10th, 11th, and 12th, it was as from the 3rd to the 6th; from the 13th 

 to the 19th it was normal, though on the last of these days it fell to zero 

 at 11 a.m. From the 20th to the 27th it was again abnormal, — positive till 

 about 9.30 a.m., then negative and then positive again, — except on the 23rd, 

 when the lowest reading obtained was zero. After the 27th the readings were 

 all positive. 



It will be observed that all the negative readings were obtained when the 



