METEOROLOGICAL OBSERVATIONS ON BEN NEVIS. 59 
agreeing with the small amount of sunshine registered by the sunshine 
recorder. 
On Ben Nevis the following phenomena were observed :— 
Auroras :—September 20 ; November 1 ; December 25. 
St. Elmo’s Fire :—January 3, 6, 27, 28; February 27 ; March 5, 17, - 
27; May 16 ; July 23; August 3, 15, 18, 19, 20, 21 ; September 2, 10; 
October 5, 15, 16. 
Thunder and Lightning :—July 2; August 15, 19, 20, 24. 
Thunder only :—June 25, 29. 
Lightning only :—January 26, 27 ; February 20, 25; March 5, 30; 
June 29. 
Solar Halos :—April 19 (with Mock Suns) ; June 29. 
Lunar Halos :—January 5, 8 ; February 9 ; July 31 ; December 1. 
In November 1903 the Deutsche Seewarte at Hamburg, Germany, 
applied to the Directors for daily telegrams from the Fort William and 
Ben Nevis Observatories, witha view to their use, along with similar 
data from Continental high-level stations, in aiding the preparation for 
the daily forecasts for the German Empire. These telegrams have 
accordingly been sent for several months past, and appear regularly in 
the Daily Weather Report issued by the ‘Seewarte.’ 
Copies of the Ben Nevis observations have also been"sent on applica- 
tion to Dr. Hergesell, to be used in connection with the International 
Aéronautical Investigation. 
The discussion of the Ben Nevis observations on the lines indicated 
in our last year’s Report has been continued. The chief subjects taken 
up by Dr. Buchan have been a continuation of the inquiry into the 
relations of temperature and pressure at the two Observatories, more 
particularly in regard to the great movements of the atmosphere grouped 
under the cyclone and the anticyclone. The observations have been 
sorted out into the following four classes :— 
1. The data for the mean hourly differences of the sea-level pressures 
and temperatures for the months, including all types of weather, with the 
exception of those days on which strong winds occurred, which rendered 
the barometric readings untrustworthy owing to the pumping of the 
mercury. 
2. The second class included all those days during the fourteen years 
on which the difference of temperature, on the mean of the whole day, was 
12°-0 or less. 
3. The third class included those days on which the difference of 
temperature was 18°-0 or upwards. 
4, All the other days were grouped under this class on which the 
difference of temperature lay between 12°-0 and 18°-0, showing thus the 
results for the days when the temperature differences were virtually about 
the average. 
This fourth class has been added to the investigation since the 
meeting of the British Association at Southport. Further, the averages 
for these four different classes have been now calculated from all the 
available observations made at the two Observatories from August 1900 
to December 1903. 
The broad results are these :—(1) When the difference of the mean 
