86 
NATURE NOTES 
are very good reasons why Winter should be considered as 
commencing on November 21 (inclusive). As March is often 
a cold month (notwithstanding its display of crocuses and 
daffodils), I would extend the dominion of Winter to the Vernal 
Equinox, and allow it to end on March 20 (inclusive). The 
counter-argument that many plants are now in bloom is of little 
force, as flowers, both wild and cultivated, may often be seen 
even in January. It should be remembered that Winter has its 
blossoms as well as Spring. The primrose, pilewort, lesser peri- 
winkle, violet, dog-mercury, crocus, daffodil, &c., are all hyemal 
as well as vernal. When we examine the temperature statistics 
we find that the minimum temperature of the nights of March 
is, on the average, not very much higher than that of the nights 
of January. At Ross (Herefordshire), for instance, the difference 
is but I '5°, and in the Scilly Islands it is only 0-2°, and this is 
in favour of January. Indeed, taking night and day together, 
March in these islands is but 0‘3° warmer than the first month 
of the year. Another argument in this connection is that at the 
Vernal Equinox the forest trees, with the exception, perhaps, of 
the more sheltered larches, have scarcely begun to put forth their 
leaves, and the general aspect of the woodlands is then decidedly 
hyemal. If this arrangement were adopted, the Winter season 
would consist of 120 days. The transitional season of Spring 
would have 62 days, and that of Autumn 58. In Leap-year, 
Winter, of course, would have 121. 
Having regard to the long duration of the Summer and 
Winter seasons, I would divide them both into “Early,” “ Mid,” 
and “ Late.” In Summer the first division would end with June 
30, and would have 40 days ; the other two would have respec- 
tively 43 and 42. In Winter the first division would terminate 
on December 31, and would consist of 41 days; the remaining 
two would contain respectively 40 and 39. 
My Midsummer day is July 23 (64-1° at Greenwich). It is 
a well marked day, for the mean temperature afterwards falls 
considerably on the average of fifty years, and does not again 
reach 64° or upwards till August 7. 
Further remarks as to November are desirable. .\t about 
5 p.rn. of November 28, 1890, the temperature at Beddington, 
Surrey, was almost unbelievably low, viz., 2’3°. I think that 
wood was heard to crack with the cold. Singularly enough, on 
the 29th the minimum at Madrid was 9'5” — positively the lowest 
on record during the period from i860 to 1894. Ross, only 
last November, the minimum of the 24th was ii-i”, while the 
maximum did not exceed 30°! Quite early in the meteorological 
Winter (December 4 in 1879) the lowest temperature (23° below 
zero) ever recorded in the British Islands took place at Black- 
adder in Berwickshire; and although this value is looked upon 
with much suspicion, there is no doubt that the weather then was 
excessively severe. It sometimes happens that the minimum 
temperature of November is also the minimum of the Winter. 
