FOE, SOUTH-WEST HEETFOEDSHIEE. 
133 
It is scarcely necessary to mention that this calendar does not 
include the results of all the observations made at Watford. 
I have excluded all of less than four years. With a larger mini¬ 
mum number of years than this I could not have included any of 
the phenomena added in the revised list of the Eoyal Meteorological 
Society, which came into operation in 1883. I have also excluded 
a few abnormally early dates, and a few exceptionally late ones, 
the former, for instance, not showing the true date of general 
first-flowering, but probably only of a single precocious plant, 
and the latter being due to omission of notice of first-flowering 
at the actual time of its occurrence. The phenomena have also 
been selected with a view to facilitate the taking of means of 
groups of observations of equal numbers, the seventy-two being 
divisible into nine groups of eight each. 
If the mean dates of all the phenomena which have been observed 
both at Watford and Odsey* are compared, it will be found that they 
are on the average about half a day later at Odsey than at Watford. 
A rather greater difference than this might have been expected. 
The number of observations from which these means are deduced 
being small—relating only to twenty-seven phenomena—the com¬ 
parison cannot be considered thoroughly satisfactory. 
The phenomena observed both at Watford and Marlborough, and 
at Watford and Cambridge, are much more numerous. A com¬ 
parison of the mean of all observations taken both at Watford and 
Marlborough shows the phenology of Watford to be three days 
later than that of Marlborough. A similar comparison of the 
observations at Watford with those at Swaffham Bulbeck near 
Cambridge, during the years 1820 to 1831, shows the phenomena 
to be on the average four days later near Cambridge than in the 
neighbourhood of Watford. Both those results are what might be 
expected from the situation of Watford, about midway between 
Marlborough to the south-west and Cambridge to the north-east; 
but too much reliance should not be placed on the results of obser¬ 
vations taken in different series of years, or even in the same series 
if the phenomena are not all observed in every year of the series. 
The greater the number of observations, the less is the liability to 
error, and that the result is as stated is doubtless due rather to the 
large number from which the mean is deduced tending to mask 
errors, than to the exactitude of the individual observations. 
* See Mr. Fordham’s Calendar in ‘ Trans. Herts Nat. Hist. Soc.,’ Yol. 1Y, 
pp. 194, 195. 
