Dec. 3] PROCEEDINGS. [1889.zhwan," to increase fromsmallbeginnings,as growing hair,"" soft,"
" weak," " to withdraw andthenincrease" ; vid supra.
zhwan,"timidly,""fearful,""cowardly";can,idru,"to turn back;"
gam, " to submit."
zhwan,"soft,""delicate,""weak," "tender," "ductile,""pliable,"
"yielding,""limber,""lithe,""nofixed principles,""infirmof
purpose,"" to stretch " ; vid. supra.
zhwan, " the crawling or wriggling of worms " : vid. supra.
zhwan, "a variety of opaque, whitishquartz, likemassivechalce
dony, withpiecesof carnelian interspersedin it." (na)za-gin,
uknti,a similar valuedstone;gin,Sa uknt elli, "of bright uknu
stone." Cp. also za, na, and dig',"stone."
I have thought it worth while to follow thisparticularsound
rightthroughthe Chinese lexicon,in order to give an example of the
highly artificial uniformity to which the old language has been
reducedin the Mandarin dialect; as well as to make it clear to all
whowill have the patience to look through these drylists, that
Chinesevocablesin their moderndisguiseare still susceptible, and
in most instanceswithout forcing, of comparison with the non-
Semitictermswhichwe find in the Assyrian syllabaries. It looks
very much as if "the pretended language of Accad" were the
forerunnerof the genuine languageof Peking.
The lexicon, as we have seen, invariablyrefers the modern
MandarinJ to an older N. A comparison of the Chinese dialects,
however,suggeststhatthe forms withinitialN are rather by-forms
which co-existed side by side withthe J (G) forms. Take, for
instance,the term ya, " tooth," which-presupposesa primitive ga,
and accordinglyappears in the dialects of Canton, Amoy, and
Shanghai, as nga, ga, nga, respectively. It is natural to compare
this term with/, ni, " to cut teeth in old age " ; a character which
in those dialectsis pronounced ngei,ge, and ni. If we find it
hard to believe in this case that the Mandarin i or ni is older
than the stronger forms of the more conservative dialects, why
should we suppose that nin is older thanjan? Prof. Douglas
informsme that nin is Japanese iorjdn,"man";but the Accadian
ni, nin, " man," " lord," which existed side by side with gin,
"man," show thatwe cannot safelypronounce offhand thatthe
one form is older than the other. This is one of those facts
which ratherincline me to suppose thatthe Accadian language,
67 G