597 
of Edinburgh, Session 1874-75. 
But it cannot be denied that a strong impetus to this science 
has latterly been given, arising partly from a more extended 
knowledge of the forms of speech since Europeans began to study 
the cognate languages of the East. Comparative philology has 
thus assumed a more definite shape within the last fifty years, as 
for instance in the law of sound-change first pointed out by Bask, 
and afterwards confirmed and extended by Grimm. 
Other phenomena of change have still more recently be6n made 
prominent, and to some of these I now wish shortly to direct 
attention. 
An opinion prevails among several eminent philologians that 
the letter and sound of l did not originally occur in the Aryan 
family to which our chief European languages belong. Its intro- 
duction, if it is not original, is certainly not recent, for it would be 
difficult to maintain that it has not existed for several thousand 
years, as it plays so conspicuous a part in the Homeric writings. 
But it appears that the Zend language — that is, the old Persian or 
Bactrian — had no such letter as Z, and that European words which 
have that sound have frequently Zend forms where r supplies the 
place of Z. It is said also that in the oldest Indian writings the 
same peculiarity appears, though the Z has been freely introduced 
into the later Sanscrit. 
Be this as it may, it must be admitted that there is a great 
affinity between the smooth and the rough liquids, Z and r, and 
that they are frequently interchangeable. We see much of this in 
Greek and Latin, and it is not easy to say that either of the two 
languages shows a preference for one of those letters over the 
other. Let us take some plain and undoubted examples : — 
Aeipiov, G. ; = lilium, L. ; paws, G., = Aa/cos, G., a ragged garment ; 
in connection with which it has been specially observed tha^ the 
Cretan form of Doric frequently confounded p and A. The ter- 
minations -pos, G., and -lus, L., seem cognate, as in rpopepos and 
tremulus. In Latin itself we have two terminational forms that 
seem identical — alis and aris — the use of which seems in a great 
measure determined by euphony, in this way, that where Z occurs 
in the radical word, the termination -aris is used for the sake of 
variety ; and when r occurs in the radical, -alis is used. Thus 
from jpopulus comes popular is ; and from naturu , naturalis. 
