126 
Magnetic Condition of the Earth, and in the Distribution 
of Temperature on its Surface.” * I showed that different 
years have temperature periods of different lengths ranging 
from about 23J to 31 days, the minimum values occurring 
in years of minimum solar spot frequency, and the maximum 
in years when the solar spots are most numerous ; the, ap- 
proximate mean value as derived from observations made 
at various stations in Europe, Asia, and North America, 
during the years 1833 to 1859, being 27'4 days; and it was 
also shown that although in any particular year observations 
at places scattered widely over the three Continents, gave 
the same value of the temperature period yet the epochs of 
maximum and minimum often differed very considerably, 
thus indicating that the variation was not due to any direct 
action of the sun, the effects of which would have occurred 
simultaneously at all the stations. The adoption of any fixed 
value of a temperature period whether that of the sun’s 
geocentric rotation on his axis, or of the moon’s synodical 
revolution round the earth, or any assumed period, did not 
satisfactorily represent observations made in different years, 
and gave much greater probable errors than a variable tem- 
perature period whose changing values followed the course 
of the curve of sun spot frequency; and thus while the 
results given by the Astronomer Eoyal in the ¥ Astrono- 
mische Nachrichten,” No. 934, do not support the hypothesis 
of a temperature period depending on the sun’s rotation on 
his axis, they will, nevertheless, on careful examination be 
found to indicate changes of temperature occurring in a 
period which was gradually diminishing from the com- 
* Proceedings, Vol. III., p. 251 — 260. 
