the highest offices of the empire, instituted a comparison between the exemplars of Lu and Ch'i, with a view to determine the true text. The result of his labors appeared in twenty-one Books, which are mentioned in Liu Hsin's catalogue. They were known as the Lun of prince Chang [1], and commanded general approbation. To Chang Yu is commonly ascribed the ejecting from the Classic the two additional books which the Ch'i exemplar contained, but Ma Twan-lin prefers to rest that circumstance on the authority of the old Lun, which we have seen was without them [2]. If we had the two Books, we might find sufficient reason from their contents to discredit them. That may have been sufficient for Chang Yu to condemn them as he did, but we can hardly supposed that he did not have before him the old Lun, which had come to light about a century before he published his work.

7. In the course of the second century, a new edition of the Analects, with a commentary, was published by one of the greatest scholars which China has ever produced, Chang Hsuan, known also as Chang K'ang-ch'ang [3]. He died in the reign of the emperor Hsien (A.D. 190-220) [4] at the age of 74, and the amount of his labors on the ancient classical literature is almost incredible. While he adopted the Lu Lun as the received text of his time, he compared it minutely with those of Ch'i and the old exemplar. In the last section f this chapter will be found a list of the readings in his commentary different from those which are now acknowledged in deference to the authority of Chu Hsi, of the Sung dynasty. They are not many, and their importance is but trifling.

8. On the whole, the above statements will satisfy the reader of the care with which the text of the Lun Yu was fixed during the dynasty of Han.

SECTION II.
AT WHAT TIME, AND BY WHOM, THE ANALECTS WERE WRITTEN; THEIR PLAN; AND AUTHENTICITY.

1. At the commencement of the notes upon the first Book, under the heading, 'The Title of the Work,' I have given the received account of its authorship, which precedes the catalogue

1 張侯論.
2 文獻通考, Bk. clxxxiv. p. 3.
3 鄭玄, 字康成.
4 孝獻皇帝.

of Liu Hsin. According to that, the Analects were compiled by the disciples if Confucius coming together after his death, and digesting the memorials of his discourses and conversations which they had severally preserved. But this cannot be true. We may believe, indeed, that many of the disciples put on record conversations which they had had with their master, and notes about his manners and incidents of his life, and that these have been incorporated with the Work which we have, but that Work must have taken its present form at a period somewhat later.

In Book VIII, chapters iii iv, we have some notices of the last days of Tsang Shan, and are told that he was visited on his death-bed by the officer Mang Ching. Now Ching was the posthumous title of Chung- sun Chieh [1], and we find him alive (Li Chi, II. Pt. ii. 2) after the death of duke Tao of Lu [2], which took place B.C. 431, about fifty years after the death of Confucius.

Again, Book XIX is all occupied with the sayings of the disciples. Confucius personally does not appear in it. Parts of it, as chapters iii, xii, and xviii, carry us down to a time when the disciples had schools and followers of their own, and were accustomed to sustain their teachings by referring to the lessons which they had learned from the sage.

Thirdly, there is the second chapter of Book XI, the second paragraph of which is evidently a note by the compilers of the Work, enumerating ten of the principal disciples, and classifying them according to their distinguishing characteristics. We can hardly suppose it to have been written while any of the ten were alive. But there is among them the name of Tsze-hsia, who lived to the age of about a hundred. We find him, B.C. 407, three-quarters of a century after the death of Confucius, at the court of Wei, to the prince of which he is reported to have presented some of the Classical Books [3].

2. We cannot therefore accept the above account of the origin of the Analects,-- that they were compiled by the disciples of Confucius. Much more likely is the view that we owe the work to their disciples. In the note on I. ii. I, a peculiarity is pointed out in the use of the surnames of Yew Zo and Tsang Shan, which

1 See Chu Hsi's commentary, in loc. -- 孟敬子, 魯大夫, 仲孫氏, 名捷.
2 悼公.
3 晉魏斯受經於卜子夏; see the 厤代統紀表, Bk. i. p. 77.

has made some Chinese critics attribute the compilation to their followers. But this conclusion does not stand investigation. Others have assigned different portions to different schools. Thus, Book V is given to the disciples of Tsze-kung; Book XI, to those of Min Tsze-ch'ien; Book XIV, to Yuan Hsien; and Book XVI has been supposed to be interpolated from the Analects of Ch'i. Even if we were to acquiesce in these decisions, we should have accounted only for a small part of the Work. It is best to rest in the general conclusion, that it was compiled by the disciples of the disciples of the sage, making free use of the written memorials concerning him which they had received, and the oral statements which they had heard, from their several masters. And we shall not be far wrong, if we determine its date as about the end of the fourth, or the beginning of the fifth century before Christ.

3. In the critical work on the Four Books, called 'Record of Remarks in the village of Yung [1],' it is observed, 'The Analects, in my opinion, were made by the disciples, just like a record of remarks. There they were recorded, and afterwards came a first-rate hand, who gave them the beautiful literary finish which we now witness, so that there is not a character which does not have its own indispensable place [2].' We have seen that the first of these statements contains only a small amount of truth with regard to the materials of the Analects, nor can we receive the second. If one hand or one mind had digested the materials provided by many, the arrangement and the style of the work would have been different. We should not have had the same remark appearing in several Books, with little variation, and sometimes with none at all. Nor can we account on this supposition for such fragments as the last chapters of the ninth, tenth, and sixteenth Books, and many others. No definite plan has been kept in view throughout. A degree of unity appears to belong to some books more than others, and in general to the first ten more than to those which follow, but there is no progress of thought or illustration of subject from Book to Book. And even in those where the chapters have

1 榕村語錄,-- 榕村, 'the village of Yung,' is, I conceive, the writer's nom de plume.
2 論語想是門弟子, 如語錄一般, 記在那裡, 後來有一高手, 鍊成文理這樣少, 下字無一不 渾.

a common subject, they are thrown together at random more than on any plan.

4. We cannot tell when the Work was first called the Lun Yu [1]. The evidence in the preceding section is sufficient to prove that when the Han scholars were engaged in collecting the ancient Books, it came before them, not in broken tablets, but complete, and arranged in Books or Sections, as we now have it. The Old copy was found deposited in the wall of the house which Confucius had occupied, and must have been placed there not later than B.C. 211, distant from the date which I have assigned to the compilation, not much more than a century and a half. That copy, written in the most ancient characters, was, possibly, the autograph of the compilers.

We have the Writings, or portions of the Writings, of several authors of the third and fourth centuries before Christ. Of these, in addition to 'The Great Learning,' 'The Doctrine of the Mean,' and 'The Works of Mencius,' I have looked over the Works of Hsun Ch'ing [2] of the orthodox school, of the philosophers Chwang and