inartificially presented, is not only a very important but a very
interesting field of inquiry.



Measurement of Man's Influence.

The exact measurement of the geographical and climatic changes hitherto
effected by man is impracticable, and we possess, in relation to them,
the means of only qualitative, not quantitative analysis. The fact of
such revolutions is established partly by historical evidence, partly by
analogical deduction from effects produced, in our own time, by
operations similar in character to those which must have taken place in
more or less remote ages of human action. Both sources of information
are alike defective in precision; the latter, for general reasons too
obvious to require specification; the former, because the facts to which
it bears testimony occurred before the habit or the means of rigorously
scientific observation upon any branch of physical research, and
especially upon climatic changes, existed.


UNCERTAINTY OF OUR HISTORICAL CONCLUSIONS ON ANCIENT CLIMATES.

The invention of measures of heat and of atmospheric moisture, pressure,
and precipitation, is extremely recent. Hence, ancient physicists have
left us no thermometric or barometric records, no tables of the fall,
evaporation, and flow of waters, and even no accurate maps of coast
lines and the course of rivers. Their notices of these phenomena are
almost wholly confined to excessive and exceptional instances of high or
of low temperatures, extraordinary falls of rain and snow, and unusual