________________
SEPTEMBER, 1882.]
MISCELLANEA.
265
(svarita) by our so-called grave accent (thus, yà). be viewed as having either the one character or No more suitable sign than the latter could be the other. For it would be as good as imposdevised, since the tone signified by it is in fact & sible to provide a complete set of vowel-signs, downward slide forward.
unaccented and accented, with a mark of nasality Passing now to the consonants, the first question added. Whether n or m shall be used as basis, concerns the mode of writing the aspirate mutes. and what and where the diacritical mark applied, And here, the addition of an h to the non-aspirate must be mainly a matter of arbitrary selection : is well nigh universal; Bopp's added reversed I prefer a dot above rather than below, because apostrophe-asť etc.-is hardly any longer in the dot below is already in full use as lingual use. In this there is nothing to be regretted; the mark, and because the dot above seems like a re. element by which the aspirate differs from the production of the corresponding devandgarf sign; non-aspirate may be sufficiently well signified by and further the adoption of the latter allows us h, nor does the distinction of surd and sonant in to write for a more independent anusvdra, and regard to it need to be insisted on. As to the i for an m assimilated to a following consonant mute-classes, the marking of the linguals (or by --- distinction which has a high practical whatever other name we may call the murdhanya convenience. class) with a dot beneath-thus, , d-is also nearly of the semivowels, only the palatal and labial without exception, and unobjectionable. But the call for discussion. For the latter of these, too, treatment of the palatals is a harder question, and is so generally current as representative that it embarrassed moreover by the doubt concerning may almost pass for universal; a few Germans the precise phonetic value of the sounds at & use w instead, but for no good and defensible given period. To me, C and ; (with, of course, reason. Historically best, to be sure, would be a ch and jh as aspirates) seem on the whole to be no in the English sense and having the English preferred : accented gutturals (as K 9') are more utterance. Yet the English sound is also ori. burdensome, and also interfere with the clearness ginally represented by v; and as we write both of the actual accent; nor should, on theoretical Latin vinum and French vin, recognizing the grounds, any diacritical mark be employed with 10-Bound as belonging to the earlier word and the 80 diverse values. This last reason is conclusivo v-bound to the later, we may properly enough do also against the common English use of ch and the same in the Sanskrit. For the palatal chh-in which, moreover, is involved a needless semirowel are widely used both y and j. The waste of time and labor.
latter has much in its favor, being in all respecta Of the nasals, n and m pass without question; related to i as v to u; and it is to the Germans and R for the lingual, goes by constraint of the natural sign for the sound, as is y to the analogy with t, d; as regards the two others, English and French. The choice of designation considerations of convenience must determine. has to be made in connection with that for the One of them will naturally be written ñ, because sonant palatal mute; and there is, it may fairly that sign is widely found already provided in fonts be claimed, & gain of convenience and economy of type; and, in accordance with its general value, in adopting for the two sounds j and y, rather this is best assigned to the palatal nasal. For than in taking and j, and so leaving y out of the remaining guttural is oftenest met with an we altogether. n with short horizontal line above it--which line
Among the sibilants we have only one fixed ought, by its length or otherwise, to be well point, the dental s; in regard to the other two distinguished from the macron.
usage is very fluctuating, and the prevailing In connection with the nasals may be considered
practioe not altogether to be approved. It was the representation of the anusvdra, difficult both
apparently by some mishap that at the outset sh on account of the variety of methods employed, came to be used by the English for the lingual and because, with the Hindu phonetists as well instead of the palatal sibilant, the two being as with their modern successors, there has been regarded as practically undistinguished in utterquestion as to the phonetio value of the sound : ance (for the definition of the lingual as like sh whether and how far it was a nasalization of the in shun, and the palatal as like 88 in 888sion, vowel, or a nasal element following the vowel.
though servilely copied from one grammar to Since, however, the Hindu texts in general tige another down to the latest, really means this, sinoe the same sign for all the different classes of omnes,
the sounds in the two words are precisely the and whatever their theoretic estimate of the sound,
same); the impression was thus given that the there appears to be no good reason why we should
lingual was the normal sh-sound, and the error has not do the same thing with the same unanimity: been perpetuated in a great variety of ways. There writing, for example, hansa, and allowing its tol is one wholly unobjectionable mode of correcting
