32 
language of the Semitic Assyrians, who subsequently established their 
domination over them, and of the Peruvian, Maya, and Aztec civilisa- 
tions of the Western World. The confusion of tongues and admixture 
of nations has increased our knowledge, for the decipherment of one 
decree has often served as a guide to the rest, and it is noteworthy 
that a ruler’s name and titles have always proved the first clue to these 
hieroglyphical records in all lands. The pursuit of this, the only royal 
road to learning with which I am acquainted, yields up to the patient 
philologist the formal structure of long dead languages, throws fresh 
light on the construction and affinities of kindred living members of 
the same linguistic type, and illustrates the successive mental and 
material phases in the development of language in general. Step by 
step we can trace the inception and growth from the rudest picture 
word of the graphic systems of the polysynthetic languages of the 
American Continent, the less agglutinative Accadian of Babylonia, the 
isolating Chinese, the inflected dialects of Semitic Assyria, Phoenicia, 
and Arabia. as well as for Aryan, Persian, and the more ancient 
Egyptian tongue. 
We now approach two of the most difficult problems of philology, 
—the development of roots and the origin of the different linguistic 
families or genera, to which the one thousand surviving specific 
Janguages are respctively affiliated. Grammar is the permanent 
framework of language, roots are the backbones—the parent stems of 
later word formation. To a comparatively small number of roots 
languages are ultimately reduced by analysis ; that is to say philologists 
can get no farther, but it must not be assumed that this ultimate 
solution represents the original ideas of the primitive framers of 
language, or the first sounds in which they expressed them. There 
are 500 roots in Chinese ; Renan gives the same number for Hebrew. 
Coptic has been reduced to about 700 word stems, and Max Miiller has 
recently resolved the 450 Sanscrit roots to 120 concepts, or mother 
ideas, expressing actions and being. Many of these, as even Max 
Miiller admits, are capable of originating in the rhythmical cries 
accompanying concerted action, as M. Noiré has suggested. Such 
verbs Romanes well shews to be exactly those to be best fitted to 
survive as roots. He further reduces these 121 concepts to about one 
hundred. Max Miiller maintains that every idea that passed through 
the mind of India as revealed in its literature can be traced back to 
these 121 concepts, and the words that have been derived from them 
by extension of sound and meaning. He adds that there are few 
concepts in English, or Latin, or Greek, which could not be expressed 
with the words that have sprung from these Sanscrit roots, and “that 
‘every thought that ever crossed the mind of man can be traced back to 
about 121 simple concepts” (Science of Thought,” pp. 417 and 418). 
These comprise such words as to milk, to gather, to dress, to adorn, to 
bake, to sew, to weave, to hate, to think, to know, and to measure. 
Proof enough, if all other evidence were lacking, that these people 
were in a relatively civilised condition very far removed from that of 
primitive man. Surely it is obvious we must seek farther back than 
this for the first developed roots of language. 
