346 
mations, except indeed so far as they are purely interjectional 
(and language Noire admits is not founded on interjections), 
are not remembered and repeated, or repeated without being 
remembered on subsequent occasions. A boy in an excited 
state may exclaim “ Bow-de-dow-de-dow,” but on a subsequent 
similar occasion the probabilities are enormously against his 
repeating this particular sound ; it has an extremely poor 
chance of passing into a “ phonetic type.” Thus, so far as the 
* evidence afforded by what now takes place in games and dances 
is concerned, we find no confirmation of the principle laid down 
by Noire, and this implies that these incidents illustrate a 
contrary principle. But, leaving this illustration, let us simply 
take the vital point of the theory. This common activity was 
accompanied by sounds, “ and AS they recurred with every re- 
petition of the particular form of activity,” etc. But did they 
so recur ? Man sat down in company to rub two stones ; he 
exclaimed, casually and by, chance, mar. He sat down again 
next day for this purpose, and again exclaimed, as of course, 
mar. Having once said mar by accident, he subsequently 
always said it again either by accident or otherwise. After a 
few more times, the sound mar became associated in his mind 
with the idea of rubbing. Then mar became a phonetic type, 
a word, subsequently a root, lastly, the parent of a tribe of words 
all connected with the one idea of rubbing. About this last 
point there is no question ; mar is an absolute fact. It was 
the sire of the god Mars, of Ares, and of the blustering Vedic 
storm-winds, the Maruts. 
Professor Noire thus holds that man pitched upon his par- 
ticular sound, e.g ., mar , in the first instance, and then adhered 
to it ever after. Of course, his view is merely a theory ; it is 
what may have been, and therefore the only standard by which 
we can try it is that of probability. Now let x = the number 
of sounds, evidently a large number, which man might or 
could have used on the original occasion ; then the probabilities 
are x to 1 against his selecting mar. But when he had once 
used it as a mere sound on a particular occasion, are the proba- 
bilities that he would use it on a subsequent occasion increased? 
Certainly not. Nature usually exhibits a repetition with varia- 
tions, not an exact repetition. He might have said kar or tar , 
etc. Looking at the question from this standpoint, Professor 
Miiller, naturally enough, sees no reason to believe that man 
pitched upon mar in the first instance. He observes : — “ Every 
possible combination of consonants, with final r or b , was sug- 
gested ; hr, tr , chr , glr , all would have answered the purpose, 
and may have been used, for all we know, previous to the first 
beginning of articulate speech. But, as soon as mar had got 
