REPORT—1847. 
profound investigation of their differences as well as of their similarity. T:: 
rigid Indo-Germanic school has assumed, but never even attempted to;^. 
that we must reject all proofs of historical affinity not resting upon tbeida- 
tity of inflexions and formative words. Now we agree with that scbod d 
maiotuining, tlmt analogies iu the musical clement of language (if 1 naj* 
call whatever belongs to the peculiarities of intonation, and the greto* 
less prevalence of one or the other class of sounds) arc in themselves asnxA. 
elusive reasons for establishing a connexion in kind, as the varieties of 
the form of leaves, smell, and similar jiropertie# are for constituWg 
species among plants, or as analogies in the colour of hair or ol 
denying the identity of species among animals. Hut 1 can see no gniurfr 
the assumption that, where identity or atlinity fails in the grammatica 
and iheir expression, there can be no radical affinity of languages. Fort* 
on this narrow principle tliat lliotie isolating systems are desjgtiedly 
con.teiously foundeil. Now we ask (anticipating what we hope 
establish), what are (according to their own a>8umption ur adinheiM)'*' 
syllables or words of inflexions but remnants of some of the substantial f 
or words (nouns and verbs), once taken out of the then common stock #0" 
tegral words, and by a convcntioniil act stamped to l)e pronouns, prcp^l^ 
or other pai’ticlcs, which gradually dwindled into inflexional 
ask further, tliia being the case, is it not an the contrary probable, tW*? 
some fumilies arc allied both by decayed and living roots, othew <“*1 
allied l)y living oiio.s only, the contemporaries of those roots wbicb - 
wards became forms aud consecjuenlly duenyed? Should ootaii 
the roots of nouns and verbs be as good evidence of a more remote, but ^ 
original connexion and consanguiJiity, as the agrceraeiit in inflexions isall'’* 
to be for tlie nearest relation between them? Languages rebted 
tity of forms (viz, by roots, once consecrated for grammatical 
then decayed) cannot exist at all without au identity or analogy w “ 
roots. IVrsons related by a common father must have a 
father, riu-re can be no identity of grammaticized and therefore 
roots, without an historical connexion of the same languages in * 
nouns and their derivatives. But a general affinity in the rooL* pn>'J_ 
common origin and a common history anterior to that point iu the _ 
ment of a language at which the grammatical forms took their origin; » 
fore a more remote one. We lay it dowm as a demonstrated sn 
testable fact, that a near aiflnity between languages is 
an identity of structure in the inflexions and the fonuatire wo^^. , 
hies m general. We have, by a combination of research and 
study, established a method to investigate this nearest 
But why should wc despair of finding also a strictly scientific 
investigating a more remote aflinity by a comparison of the rpoW ^ 
substantial words? You have hitherto studied the natural t 
most grammatical (and therefore, I believe, voongest) languages: y«“. 
hus found .a method for understanding the latest part in the f . 
representing therefore, I suppose, the most recent period in the 
luman speech. Of course this method w ill not cany you further: 
IS the reason why you have always signally failed, whenever you w 
empted to investigate languages beyond that narrow family-coo”®’?’' 
wium you have attempted to establish an affiuity between the 
aiul a hinnatioji anterior to that individual system of forms, as wr ' 
trie iJa-«({uc language. Still we cannot proceed further in comparj” ^ 
ethnology, without investigating that pro 
«»«st therefore ask two questions: why should tLre not be an a®”’- 
