intercourse, that their inflexions were different, and their 
very vocabulary so altered that they were no longer mutually 
intelligible. That this process is now going on in many places 
we learn from travellers. The Indians on the Amazon, we 
are told, speak languages differing in an extraordinary manner, 
and varying so much that a person who has learnt to express 
himself with tolerable fluency in conversation with a certain 
tribe will with difficulty understand or be understood on 
revisiting them after the lapse of twenty or twenty-five years. 
Superstition, too (as I have said), exercises a great influence 
on the vocabulary, if not on the grammar. In some nations 
the king takes the name of some animal or object, which 
name is forthwith banished from the language, since any one 
using it would be immediately suspected of trying to bewitch 
the chief. A new noun has to be invented and thenceforward 
employed to designate the object. In others the fetish of 
the community, or the instrument of some good or evil to 
them, must no longer be called by the name it bore up to that 
period. So the greatest ingenuity has to be exercised in the 
formation of new words which shall be as different as possible 
from the old ones. It does not always happen that two 
branches of the same tribe invent the same new appellative ; 
and hence a variation which a very few years suffice to convert 
into an actual breach of continuity. 
5. To these disturbing forces we may add the occasional 
intermixture of foreign individuals. These intermixtures were 
rarer in early times ; but still there is no reason to doubt that, 
when they did occur, the presence of a few influential strangers 
had a tendency to introduce new words into the vocabulary, 
and perhaps to affect in a perceptible degree the use of pre- 
fixes, suffixes, and medial changes; or that conquerors or 
slaves would compel their subjects or masters to accept some 
of their language, and (in JuvenaPs words) make Orontes 
flow into Tiber. 
Such are the principal causes of the alteration, develop- 
ment, and decay of the forms of human speech. Nor will it 
be correct to argue that they affect vocabulary only, and not 
grammatical character ; that they quite account for the evolu- 
tion of Persian out of Pehlvi, or of Hindi out of Sanskrit, but 
cannot be adequate to explain how from one origin there could 
spring tongues so radically different as Manchu and German. 
True, the grammar of a written language is invariable in every 
direction but one. No philological circumstances could ever 
make Italians form the plural with s, or Spaniards without it. 
But that is owing to the fixity given by written, or at all 
events traditional, literature. To an early tribe, using a simple 
