288 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1907. 
study of the motions of the atmosphere. We must define it as em- 
bracing the broadest conceivable study of the atmosphere from any 
and every point of view; if we subdivide it according to the difficulty 
of the subject and the extent of our ignorance, beginning with the 
simplest and passing on to the more difficult portions, we may sub- 
divide it into descriptive climatology, optical and acoustic phenom- 
ena, thermal or thermodynamic phenomena, hydrodynamic or me- 
chanical phenomena. The two latter classes of phenomena consti- 
tute the subject-matter of the mechanics of the atmosphere and 
include all that relates to temperature, pressure, winds, cloud, fog, 
dew, rain, snow, hail, and the daily predictions of storms and 
weather. It is in the study of these phenomena that the progress 
of our science has been most conspicuous during the past century; 
the problems already partially solved involve so much of the most 
profound modern physics and mathematics that one can not refuse 
to meteorology a notable place among the most difficult branches of 
science. Fundamentally, meteorology is the mechanics of the earth’s 
atmosphere; all its other aspects are of minor importance to this and 
it is our progress in this line of research that should especially claim 
our attention. 
A general historical survey of the methods by which we have 
arrived at the present state of our knowledge of nature will show 
that meteorology has passed through the various stages of develop- 
ment that have been common to all the sciences, and that in its pres- 
ent stage of vigorous growth it already stands among those that have 
progressed the furthest. The methods of advancing our knowledge 
of nature have been the same in all ages, among all nations, and in 
almost all individual cases. One individual, or one nation, or one 
age may differ from another in its predilections for special methods, 
but in general we find everywhere analogous methods of thought and 
work, and they even succeed each other in the same order. Begin- 
ning with explorations and crude observations, man passes on to 
generalizations and inductions. If possible he frames speculations 
or working hypotheses as to the ultimate cause or the rationale of any 
phenomenon, and then tests his tentative deductions by experimenta- 
tion until the working hypothesis has been so modified as to represent 
some general law. The association of several such laws leads to the 
building up of elaborate deductive theories, not speculations in the 
popular sense of the word, but well-established systems, or methods 
of argumentation, that represent a rational and more or less profound 
knowledge of nature. Such “theories” are well exemplified by 
Gauss’s “ Theoria Motus” or Rayleigh’s “ Theory of Sound.” 
If at some epoch a man or a nation is unable to apply any one of 
the above-mentioned methods of study, then the real knowledge of 
nature stops at that point, and man waits until the development of 
