y, for there have
been instances, as in the case just quoted of Philolaus, where the results
properly belonging to the third have been anticipated in the first stage.
To the quick eye of genius one case may be like a thousand, and one
experiment, well chosen, may lead to the discovery of an absolute law.
Besides, there are great chasms in the history of science. The tradition
of generations is broken by political or ethnic earthquakes, and the work
that was nearly finished has frequently had to be done again from the
beginning, when a new surface had been formed for the growth of a new
civilization. The succession, however, of these three stages is no doubt
the natural one, and it is very properly observed in the study of every
science. The student of botany begins as a collector of plants. Taking
each plant by itself, he observes its peculiar character, its habitat, its
proper season, its popular or unscientific name. He learns to distinguish
between the roots, the stem, the leaves, the flower, the calyx, the
stamina, and pistils. He learns, so to say, the practical grammar of the
plant before he can begin to compare, to arrange, and classify.

Again, no one can enter with advantage on the third stage of any physical
science without having passed through the second. No one can study _the_
plant, no one can understand the bearing of such a work as, for instance,
Professor Schleiden's "Life of the Plant,"(11) who has not studied the
life of plants in the wonderful variety, and in the still more wonderful
order, of nature. These last and highest achievements of inductive
philosophy are possible only after the way has been cleared by previous
classification. The philosopher must command his classes like regiments
which obey the order of their general. Thus alone can the battle be fought
and truth be conquered.

After this rapid glance at the history of the other physical sciences, we
now return to our own, the science of language, in order to see whether it
really is a science, and whether it can be brought back to the standard of
the inductive sciences. We want to know whether it has passed, or is still
passing, through the three phases of physical research; whether its
progress has been systematic or desultory, whether its method has been
appropriate or not. But before we do this, we shall, I think, have to do
something else. You may have observed that I always took it for granted
that the science of language, which is b