47 
of Edinburgh, Session 1869 - 70 . 
But there are other persons in the Pantheon whose significance is 
anything but plain ; and in their case, unquestionably, recourse may 
be had with advantage to etymology, first, in the native language, 
of course, and then in the kindred languages, in some one of 
which the original form of the sacred title may have been pre- 
served. A striking example of the utility of native etymology in 
fixing the significance of the Greek mythological personages is pre- 
sented in the familiar case of the Harpies, whose whole character 
and actions, taken along with the open evidence of their Greek names 
in Hesiod, prove, beyond all doubt, that they are the impersonated 
forms of such sudden gusts and squalls of wind as come down 
fuequently on the Black Sea or the Highland lochs. But etymo- 
logy, though a safe guide in such instances, is, in less obvious 
cases, of all guides the most fallacious. And this is what my 
distinguished friend Max Muller, and some who follow in his train, 
seem at the present moment somewhat apt to forget. An etymo- 
logy, though not caught up in the arbitrary fashion of Bryant and 
Inman, but traced with the most cautious application of Grimm’s 
laws, is, after all, only a conjecture. It is a conjecture not in the 
teeth of all philological analogy. It implies a possible, or, as the 
case may be, a probable identity. But alone, and without extrinsic 
and real, as opposed to verbal indications, it affords no ground for a 
legitimate induction. Nothing is more common than accidental 
coincidences in mythological names — such as the Latin Hercules 
and the Greek Heracles — which, as scholars know, have not the 
most remote connection. Besides, even if the true etymology of 
any Greek god could be found in Sanscrit or any other language, 
the signification of the original name affords no sure clue to the 
character of the accomplished god. Our dictionaries are full of 
words whose ultimate signification has travelled so far away from 
its original, that the original meaning could supply no key to the 
modern usage, n op<£vpeos, for instance, means dark in Homer, but 
in Horace brilliant or shining. Usage alone can inform us of this 
perversion or inversion of the original meaning of words. But 
if this be true with regard to mere philology, it is much more true 
with regard to mythology. The root of a word, like the stock of 
a tree, may remain stiff enough for centuries; but the human 
imagination, when employed in the forming of myths, is a kaleido- 
