Variation of the Solar Radiation. 89 
of the Deutsche Seewarte for nearly 100 stations, for each 
ten-day period of 1903, and accompanied by normal tempe- 
ratures representing the mean for the same ten-day periods 
of many former years*. From all the stations for which. 
normal values are given, temperature departures have been 
computed here, and these departures have been averaged 
by geographical position in seven groups, comprising (1) 
North America, 20 stations; (2) Insular and South-west 
Hurope and North Africa, 18 stations; (3) North-west Hurope, 
15 stations ; (4) Central Europe, 10 stations ; (5) Huropean 
Russia, 11 stations; (6) Asiatic Russia, 8 stations; (7) High 
Altitudes in Europe, 7 stations. ‘These seven groups of re- 
sults are plotted in the accompanying chart (PI. I.) (in which, 
to avoid confusion, the lines are alternately continuous and 
dotted), and as all seem to show the same tendency in more 
or less marked degree, no hesitation was felt in taking the 
general mean, giving observations of each station equal 
weight wherever present in the sum. The general mean for 
the 89 stations forms the (heavy) eighth line of the chart, 
and a comparison is invited between its course and that of 
the ninth and lowest line representing the observations of 
solar radiation made during 1903 at Washingtont. In con- 
nexion with this comparison, the reader is reminded that 
during September an increasing transparency of the atmo- 
sphere was noted which may have had the direct effect, dis- 
cussed at a former page, of increasing terrestrial temperatures 
thereafter, and thus accounting for the observed rise of the 
mean temperature at the close of the year. If this be ad- 
mitted, the general agreement of the observed temperatures, 
with what radiation measures would lead us to expect, is 
certainly marked. 
As for the total fall of temperature indicated by the charts, 
that is seen to be a maximum for Huropean and Asiatic 
Russia, the regions most removed from the influences of the 
oceans. The fall in the general mean for the land areas of 
the North Temperate Zonet is, however, over 2° centigrade, 
or about one-third of the maximum fall possible to be caused 
* The writer is indebted to Prof. Cleveland Abbe and to Dr. W.F. RB. 
Phillips, Librarian of the U.S. Weather Bureau, for their aid in making 
accessible the publications of temperature data in possession of the 
Weather Bureau. 
+ It is to be regretted that only one day of observation with at all 
satisfactory conditions was secured between April 29 and August 24, 
and that (July 7) of less weight than the others. 
{ It would be of great interest to extend the discussion of temperature 
departures to include stations in the interior of South America, Africa, 
Australia, and Southern Asia, but the data are not now available, and 
for many regions of these continents may long be wanting. 
