132 
J. HOPKINSOH-CLIMATE OE HERTFORDSHIRE. 
The rainfall given in these tables must not, as before mentioned, 
be considered to represent the mean for a long period, for, during 
the twelve years 1887-98, the mean annual fall in Hertfordshire 
was nearly two inches below that for the sixty years 1840-99. 
During this period, deducing the mean from the records of 
1 station for 1840-49, 2 stations for 1850-59, 7 for 1860-69, 
12 for 1870-79, 18 for 1880-89, and 20 for 1890-99, the result 
is thus:— 
ins. 
ins. 
ins. 
ins. 
March 
1-65 
June 
2-05 
Sept. 
2-35 
Dec. 
2-04 
April 
1-69 
July 
2-48 
Oct. 
2-96 
Jan. 
2-22 
May 
2-06 
Aug. 
2-44 
Nov. 
2-56 
Feb. 
1-65 
Spring 
5-40 
Summer 6-97 
Autumn 
1 
CO 
— J 1 
Winter 5-91 
The total for the year is 26-15 inches. The greatest rainfall in 
any year during this period was 37-59 ins. in 1852, and the least 
was 17*67 ins. in 1854. 
The last two tables give the results of certain other observations 
which have been recorded at Berkhamsted and Bennington. The 
temperature of the soil at Bennington, at one foot beneath the 
surface, has been taken only during the eight years 1891-98, and 
at two feet beneath the surface for the two years 1897-98, so that 
the records are not strictly comparable with those at Berkhamsted, 
which have been taken for the whole period of twelve years. Nor 
are the records of sunshine, for two reasons. They have only been 
taken at Bennington for the ten years 1889-98, and the record is 
that of a Jordan photographic sunshine-recorder. Not only does 
this give results which differ from those given by a Campbell- 
Stokes burning-glass, but the constancy of the actinic sensitiveness 
of the recording paper cannot be relied upon. The sensitiveness of 
the paper decreases the longer it is kept, and may (in fact does) 
vary from time to time as provided. With the Campbell-Stokes 
recorder, on the other hand, it is only necessary to see that the 
focus is accurate and to keep the surface of the glass clean and 
bright. It does not therefore necessarily follow that the duration 
of bright sunshine at Bennington is nearly three-quarters of an 
hour per diem longer at Bennington than it is at Berkhamsted, the 
record for which place of very nearly four hours per diem we may 
assume to be the more correct. However this may be, these tables 
show that while the sky is clear at 9 a.m. for about the same 
number of days in the year at both places, it is then completely 
overcast for 20 more days in the year on the average at Bennington 
than it is at Berkhamsted, a fact difficult to account for. 
The other results in these tables accord very well, except that 
the barometric pressure is higher at Bennington than at Berk¬ 
hamsted. The difference of one-hundredth of an inch is, however, 
very small. We cannot look for the cause in the slightly greater 
dryness of the air at Bennington, due to the more exposed situation 
and drier soil, for the presence of moisture in the air increases its 
pressure, but we may trace a result in the smaller rainfall there. 
