297 
rules, they are worth attending to and testing ; because they are 
the result of many years observations both of scientific and 
unscientific men; and looking again at Sir R. Christison’s 
remark, limiting the time to be noted to the last week in 
October and the first of November, it is curious to notice how 
almost all weather profits profess to attach great importance to 
such weeks, The idea is at least as old as Aratus who gave 
weather forecasts in the third century before Christ ; he said 
“Tt deeply imports 
To mark the last four days of the dying, 
And first four of the nascent month—of meeting months 
They join the edges when most changeable is the atmosphere 
Wanting eight nights the mild rays of the moon— 
Such notes, collated with the calendar 
Shall furnish solid forecast of the sky.” 
Phenomena, 1146—1152. 
Poste’s translation. 
Of course he was speaking of lunar months, but regardless 
of this, weather observers and weather prophets have repeated 
his rules and applied them to calendar months, with which they 
cannot possibly have any connection.* 
I have spoken at some length about the autumnal cold, 
followed by a mild winter ; not only because last winter was such a 
marked example of it, but also because we have been passing 
through a cold Halloween and Martinmas of more than usual 
length, and it may be interesting to note whether the same 
result will follow this year, but 1 am afraid we must at present 
content ourselves with hoping only. But before I quite leave 
that part of my subject I must ask you not to suppose for a 
moment that I consider the mild winter of 1896 as the result of, 
* On this subject of weather forecasts, and their connection 
with special days, see an excellent paper by our late President, in 
Vol. I1., page 161, of our transactions entitled, “St. Swithin and 
other weather Saints.” 
