AGRICULTURAL METEOROLOGY. 135 
causes which belong to the first three categories, we should always ob- 
serve identical plants, in the same localities, and with similar soil, age, 
&ec.; the differences existing in the periods of flowering, when we have 
exhausted all other equal circumstances, may be attributed to meteoro- 
logical causes. Supposing that it is the same observer who follows the 
development of the plants, and takes care to register their natural 
epochs, numerous observers would see without doubt in different ways, 
one having a propensity to mark the epoch of flowering more slowly 
than another. This difference, which may be named “the personal equa- 
tion of the observer,” will be a new cause of error, of which it is neces- 
sary to take account. 
I assume, therefore, that all the precautions have been taken to render 
as equal as possible all the causes in general, other than meteorological, 
which may make a difference in the phenomena of vegetation. ‘These 
*conditions are not very difficult to fulfill if we observe the same plants, 
in identical exposures and localities, during many consecutive years, in 
order to eliminate by repeated observations the results of casual causes; 
we should also take care to register the meteorological state of the air, 
which action is to be determined. 
Supposing that we have registered carefully each year the flowering 
periods of common and unmistakable plants, such as the lilac, (Syringa 
vulgaris,) after a series of years we may take the average of the dates 
we have preserved, and this average will give the flowering epoch of 
the plant. This epoch is not, however, so well fixed that it may not be 
affected by ulterior observations, and we may, therefore, for the present 
assign a probable error. Wemay also determine the probable digression 
of an ulterior observation of the date, for example, at which the lilac 
will flower in the following year... 
The table which Quetelet gives* throws more light on the subject. 
We find there the dates at which have blossomed for fifteen years the 
most common of plants, viz: the lilac, the syringa, and the faux ébénier, 
and also the average dates of the digressions observed each year. 
The celebrated botanist, Martius, remarks that the movements of the 
sap, studied by the aid of periodical phenomena, and placed in connec. 
tion with the meteorolegical derangements, throw a new light upon the 
history of ceils and vessels; the chemical signification of the external 
development we find more clear; the history of vegetable chemistry 
consolidates the principles of agronomy. » 
The average date of flowering may be fixed at the 1st of May for the 
lilac, the 25th of the same month for the syringa, and the 2d of June 
for the faux ébénier. We see further that the probable error of these 
dates is nearly the same, and does not extend beyond a day and a half. 
The probable error is naturally greater for an isolated observation ; it is 
of five or six days; that is to say, that, for the faux ébénier, for example, 
it is an equal chance that the flowering will take place the 2d of June, 
and that the delay or advance will not exceed five or six days. Although 
the three plants indicated bloom at different epochs, we see consequently 
that they have nearly the same delay and the same advance for each 
year, 
ACTION OF FROST UPON PLANTS. 
Theory of frost.—From a series of experiments on the “Heat of An- 
imals and Vegetables,” made in 1766, John Hunter drew the following 
conclusions: “That plants, when in a state of actual vegetation, or even 
in such astate as to be capable of vegetating under certain circumstances, 
* Theorie des Probabilities. Bruxelles, 1853, pp. 68-70. 
