338 
president’s address. 
Table I. shows the absolute highest monthly and annual shade 
temperature during the period. These are made (as are all the 
temperature observations referred to in this paper) from 
thermometers by Negretti and Zambra, exposed in a Stevenson 
screen, well situated to obtain a free current of air passing 
through it, but shaded from the sun’s rays, and screened from 
radiation by night. It will be noticed that the highest reading 
recorded during the period was 90 J in August, 1893, and that in 
no summer during the period did the thermometer fail to reach 80°. 
Table II. shows the lowest night temperature in the screen, and 
must not be confounded with the readings obtained from exposed 
thermometers placed upon the grass where much lower readings 
on clear nights are sometimes obtained. The minimum of the 
period was in February, 1895, when the thermometer fell to 4°, 
consequently, taking Tables I. and II. together, a range of 
temperature of 86° was recorded in the ten years. The extreme 
of cold in the winter is much more marked than the extreme of 
heat in the summer, for whereas the day readings were from 80° 
to 90° or a difference of 10° only, in winter the coldest night of 
the season has differed as much as 20.2°, i.e., 4° in the winter of 
1894 — 5, and 24.2° in the following winter. There is no doubt 
that some of the winters during the period were unusually 
severe, in fact 20° of frost and upwards were registered in all 
but two of them, and in five of them the thermometer fell below 
10°. The last two winters of the period (1896 and 1897) were 
by far the mildest, and gave comparatively little frost. 
Table III. gives the mean temperature of each month and year 
during the period (i.e., the means of maximum and minimum). 
The most noteworthy feature of this table is the excess in 
temperature of the summer months of the last five years over the 
first five years, and, if we except 1895, the same remark applies 
to the spring months. The yearly temperature also shows 
a similar tendency, the mean of the first five years being 47.5°, 
and that of the last five being 49.2° or a difference of 1.7°. 
Table IY. shows the monthly and annual rainfall of the ten 
years, as collected in a five inch “ Symons ” rain-gauge, whose 
