then about three thousand compound characters, and perhaps one thousand simple ones, are all for which we have yet been able to account. These however form less than a sixth of the language, even when we estimate the characters at thirty thousand. Yet it must occur to the reflective mind, that in order toapply with any degree of precision, the principles on which the language is said to be formed, the component parts of all the characters must be clearly ascertained. This can be done with ease when the compound consists of only two elements, as must have been the case with nearly all in the infancy of the language; and it may be done with almost equal certainty when a character is compounded of three without two of them forming a previous compound. But here the research must end: the attempt to trace the meaning of a character from its various component parts, when all of these beside one may have previously found a new character expressive of another idea, can lead to nothing but whim or disappointment. It is therefore of some importance that we endeavour to trace the mode in which the rest of the characters are formed, as without this we must remain in the dark respecting the formation of full five-sixths of the Chinese language. Before we enter upon this subject, it may however be proper to enquire strictly into the ACTUAL NUMBER of the characters which compose the language, particularly as the idea of its containing sixty or seventy thousand, still occasionally appears before the public. NUMBER OF THE CHARACTERS. It will not be denied that in an ancient and fixed language, the number of words which it contains may be obtained with pretty considerable cer tainty from the dictionaries in use, particularly when these have been improved in succession to any considerable extent. Now the Imperial Dictiona ry in thirty-two volumes, is at least the eighth of respectable authority which has been compiled of the Chinese characters. Six are enumerated in the preface to that dictionary as having preceded it, and seven are quoted in the body of the work. The most ancient of these, and perhaps the most ancient in the world, was compiled, according to that preface, under the Han dynasty, probably by Woo-tee (Vu-ti) the third emperor of that house, a great encou rager of literature, who is said to have ascended the throne about 140 years before the Christian Æra, and to have restored the five classical books which the Tyrant Shee (Xi) had ordered to be destroyed some years before. This was followed by five others in the space of about sixteen hundred years, and these by the Ching-tsee-toong, which was published about the close of the seventeenth century, a few years before the Imperial Dictionary. In the 49th year of his reign (A. D. 1710) the emperor Kang-kee commanded certain of his chief mandarines to collect these dictionaries, and, from them and other works, to compile one which might form a standard both for the characters and the sounds. In pursuance of this command, the principal men in China for learning, to the amount of nearly a hundred, examined the dictionaries extant, and making the Ching-tsee-toong their model for arranging the characters, added about six thousand not contained in that dictionary; but distinguished them under every element or key by prefixing the character, tsung, “added.” Beside these, they from various works collected nearly 1700 characters, which had never yet found a place in any dictionary; and lastly about 4000 more, part of which they describe as having no name, and the rest as having neither name nor meaning. These they did not admit into the body of the work, even among those added; but placed at the end in two small volumes; one of which contains 84 leaves or double duodecimo pages, and the other 54. After a scrutiny so severe as to include not merely the obsolete characters, and the irregular forms of others to which caprice or inadvertence might have given birth, but many without name or meaning,-a scrutiny made too by nearly a hundred persons advanced to rank on account of their acquaintance with their own language, it is not likely that many characters should be overlooked. In the Imperial Dictionary therefore, we may be supposed to have the sum total of the characters given; in order to ascertain which with precision, every page has been repeatedly examined, and, mistakes excepted, the number is found to stand thus: The characters in the body of the work, Added, principally obsolete and incorrect forms of others, Characters without name or meaning, 31,214 It is obvious however, that these forty-three thousand characters do not express an equal number of different ideas. Several thousand of them have no meaning affixed to them in the Imperial Dictionary: this is the case with nearly the whole of the last class; and of the 6423 added in the body of the work under the different elements, by far the greater part being either obsolete or incorrect forms of the other characters, to them the reader is of course referred for their meaning: so that of the twelve thousand which the diligent researches of the learned employed in compiling the Imperial Dictionary, have added to those contained in the preceding ones, scarcely a third of them express new ideas, or have even an explanation affixed to them. But beside these, a considerable number of the 31,214 characters adopted. from the former dictionaries, have no meaning affixed to them; but are merely given as obsolete, or current but incorrect forms of other characters, to which the compilers of the dictionary have referred the reader for their meaning. In a volume of the dictionary containing 1004 characters, the writer found 115 thus referred to other characters. The significant characters of the language, therefore, including names of every description, can scarcely be estimated at more than Thirty Thousand. Of these Thirty Thousand however, we have not been able, by including both the elements and those formed by two elements united, to account for a sixth. The grand enquiry therefore still remains to be answered: How are the rest formed? As they cannot be formed immediately from the elements, are they formed from certain primitive characters like the roots in Greek, or the dhatoos in Sungskrit? If so, what are these primitive characters, and by what method are the derivatives formed from them? This has for several years occupied the mind of the writer of this dissertation, and he is ready to hope, that he has at length succeeded in tracing the manner in which nearly the whole of the Chinese characters are formed from each other. To some it may not be uninteresting to learn by what process he has been led to this conclusion. OF THE PRIMITIVES. That such Primitives really exist as occupy the middle space between the elements and the great mass of the characters, and, like the Greek primitives of the Sungskrit dhatoos, form the bulk of the language by associating to themselves certain of the elements, was long suspected by the writer. This idea was strengthened by his observing in a manuscript Latin-Chinese Dictionary which classed the characters according to their names, that in numerous. instances, one character was the root of ten or twelve others, each of which was formed from it by the addition of a single element; thus the addition of the element for a hand to a primitive, formed one character; that being changed for the element denoting the head, another character was formed from the same root; by the change of that for fire, a third; and of that for the element denoting water, a fourth. It further appeared that the characters thus formed from the same primitive by merely adding one element, generally took the name of the primitive with some slight variation. This so struck him that he examined the dictionary from beginning to end, noting down each primitive as it occurred, and referring thereto all the characters formed from it by the addition of one element: and he at length found, with astonishment and pleasure, that all the characters of this dictionary, about nine thousand, were formed from eight hundred and sixty-two characters, by the addition of only one element. Fearing however to be mistaken in a fact that promised to throw so much light on the formation of this singular language, and reflecting that nine thousand characters bore but a small proportion to the whole mass, he by the help of his Chinese assistants, set about examining the whole of the E |