CHAPTER XI

PORCELAIN AND ITS BEGINNINGS

THE reader will have noticed that the word porcelain, which was avoided in the discussion of the earlier periods, has insensibly crept into the chapters which deal with the Sung wares. It was no longer right or proper that it should be excluded, and it is high time that our attitude on the interesting question of its origin was defined. Unfortunately, that attitude is still--and must necessarily remain--one of doubt and uncertainty, but we can at least clear away some of the existing misapprehensions on the subject.

The myth which carried back the manufacture of porcelain some eighteen centuries before our era has been definitely discredited, and the snuff bottles supposed to have been found in ancient Egyptian tombs which gave rise to the idea are now known to be of quite modern make. The more modest computation which placed the invention in the Han dynasty (206 B.C. to 220 A.D.) might have been almost as lightly dismissed had not Dr. Bushell, after disposing of the theory in his _Oriental Ceramic Art_[292] in 1899, seen fit to reverse his decision in later publications.[293]

The reasons given for this later attitude are on the surface so convincing that it is necessary to consider them in detail and to examine the authorities on which they are based. Bushell's statement runs as follows: "It is generally agreed that porcelain was first made in China, but authorities differ widely in fixing a date for its invention. The Chinese attribute its invention to the Han dynasty, when a new character _tz'ŭ_ was coined to designate, presumably, a new substance. The official memoir on 'Porcelain Administration' in the topography of Fou-liang, the first edition of which was published in 1270, says that according to local tradition the ceramic works at Hsin-p'ing (an old name of Fou-liang) were founded in the time of the Han dynasty, and had been in constant operation ever since. This is confirmed by T'ang Ying, the celebrated superintendent of the Imperial potteries, appointed in 1728, who states in his autobiography that the result of his researches shows that porcelain was first made during the Han dynasty at Ch'ang-nan (Ching-tê Chên), in the district of Fou-liang."

From this and the passages immediately following it is clear that Bushell at that time leant strongly to the Han theory, which he had previously discarded, for three reasons, which we shall now examine. The first rests on the character _tz'ŭ_. Whether the character _tz'ŭ_ was coined to designate a new substance in the Han dynasty is by no means certain. It undoubtedly appears in the Han dictionary, the _Shuo Wên_, but with the meagre definition "pottery ware,"[294] and without any further indication of its nature. The second is based on a passage in the Annals of Fou-liang, which on examination proves to contain only the general word _t'ao_ (ware) and not the character _tz'ŭ_ at all. The actual passage runs: "The manufacture of pottery (_t'ao_) at Hsin-p'ing began in the Han dynasty. Speaking generally, this pottery was strong, heavy, and coarse, being fashioned of rich clay with moisture added, after methods handed down from the ancients." The third invokes the authority of T'ang Ying, but on reference to the autobiography of this distinguished ceramist in the _Chiang hsi t'ung chih_, we again find reference only to _t'ao_ and not to _tz'ŭ_, viz. "It (_t'ao_) is not the growth of one day. Research shows that it began in the Han dynasty and was transmitted through succeeding generations. Its place (of manufacture) changed (from time to time), but it flourished at Ch'ang-nan." One obvious place for T'ang's research would be the Annals of Fou-liang, and I shrewdly suspect that his conclusions were based on the very passage quoted above, of which his words give a clear echo. But in any case, neither passage has any bearing on the origin of porcelain unless we assume that _t'ao_ is the same as _tz'ŭ_, and that both words definitely mean _porcelain_, an assumption which is not only quite unwarranted but in any case begs the whole question.

The Chinese words used at the present day for porcelain are _tz'ŭ_, _t'ao_, and _yao_, all of considerable antiquity, though their forms have undergone various changes and their meaning has been modified from time to time to keep pace with the evolution of the ware. The word _tz'ŭ_ [Chinese], as we have seen, was defined in the Han dictionary as merely "pottery ware." Its modern definition is a hard, fine-grained variety of _t'ao_, and if we add to this the quality of resonance--i.e. of emitting a musical note when struck--we have all the requirements of porcelain according to the Chinese definition. The synonym [Chinese], containing the radical [Chinese] _shih_ (a stone), which is also pronounced _tz'ŭ_, has come in the last two centuries to be used interchangeably with the older word [Chinese], in spite of the protests of eighteenth-century purists.[295]

The word _t'ao_ [Chinese] is a term even more comprehensive than our word "china." In the Han dictionary it appeared in the form [Chinese] _t'ao_ or _yao_ (previously pronounced _fou_), composed of [Chinese] _fou_ (earthenware) and the radical [Chinese] _pao_ (to wrap), and its meaning was _kiln_, and by extension the _products of the kiln_. At that time the word in its modern form was only used as a proper name.

The third character _yao_ (the Japanese _yaki_) is precisely synonymous with _t'ao_, meaning first a _kiln_ and then _wares_ of any kind. In its form [Chinese] it occurs in the Han dictionary; another form is [Chinese], which, according to a Sung writer,[296] dates from the T'ang period, and a third form [Chinese] is current in modern dictionaries.

In short, the Chinese terms are all of a general and comprehensive kind, capable of embracing pottery, stoneware, and porcelain impartially, and there is no single Chinese word which corresponds to our precise term "porcelain." Under these circumstances it is clear that no theory on the origin of porcelain can be based merely on the occurrence of any of these words in early Chinese texts. Still less can any such theory be constructed from the very promiscuous use of the word "porcelain" in European translations, and it is a thousand pities that both Julien and Bushell were not more discriminating in this matter, or that they did not always (as Julien sometimes and Professor Hirth usually did) give the Chinese character in parentheses when any reasonable doubt could exist. Had this been done we should have been spared misleading references to "two porcelain cups of the Han dynasty,"[297] and such loose writing[298] as "In the Wei dynasty (221-264 A.D.) which succeeded the Han we read of a glazed celadon ware made at Lo Yang for the use of the palace, and in the Chin dynasty (265-419) we have the first mention of blue porcelain produced at Wên-chou, in the province of Chehkiang, the progenitor of the sky-blue glazes tinted with cobalt, which afterwards became so famous." The "glazed celadon," needless to say, is purely conjectural, pottery (_t'ao_) vessels being all that is specified in the passage on which the statement is obviously based; and the "blue porcelain" is evidently no other than the _p'iao tz'ŭ_ (mentioned by the poet P'an Yo and discussed on p. 16), which is better rendered "green ware."[299]

The same kind of criticism applies to all the other references in early writers until we reach the Sui dynasty (581-617 A.D.). In the annals of this period there is a much discussed passage in which it is stated that the art of making a substance known as _liu-li_[300] had been lost in China, and that the workmen did not dare to experiment, but that one Ho Ch'ou [Chinese], a connoisseur in pictures and antiquities, succeeded in making it with green ware (_lü tz'ŭ_), and that his imitations were not distinguishable from the original substance.

To understand the full import of this passage it is necessary to explain the nature of _liu-li_, and this is fortunately made quite clear by the author of the _T'ao shuo_ in a commentary so interesting that I give it in full:

"I find that _liu-li_ comes from the countries of Huang-chih, Ssŭ-t'iao and Jih-nan.[301] That produced in Ta-ch'in (the eastern provinces of the Roman Empire) is in ten colours--pink, white, black, yellow, blue, green, deep purple, deep blue (or green), red, and brown. _Liu-li_ was originally a natural substance.[302] Yen Shih-ku,[303] commenting in the Annals of the Han Dynasty, says, 'At the present time they commonly use molten stones, adding a number of chemicals and then pouring the substance (into moulds) and forming it; but it is unsubstantial, brittle, and not a successful casting.' In the Northern Wei dynasty, in the reign of T'ai Wu (424-451 A.D.), a man of the Ta Yüeh-chih[304] who came to trade at the capital, said he could make _liu-li_ by melting stones. Eventually he collected the ore and made it (_liu-li_), and the finished article surpassed the original in its brilliance and colour. The method has been handed down to the present day, and it was probably only an accidental intermission which occurred in the Sui dynasty. But the Chinese castings are brittle in substance, and when hot wine is poured into them they fly to pieces in the hand. What a pity the Yüeh-chih method has been handed down instead of Ch'ou's!"

The allusions to melting stones, casting, etc., in this passage leave no doubt that the _liu-li_, as made in China, was a kind of glass, imitating a natural stone.[305] It is, in fact, usually translated in the dictionaries as "opaque glass," and in connection with pottery it has the sense of glaze--e.g. _liu li wa_ "glazed pottery."

We can now return to Ho Ch'ou, who "took green ware and made _liu-li_."[306] It has been thought that what he made must have been a kind of porcelain, but there is no indication of any such achievement, for though it is possible to make an artificial porcelain with glass as a constituent, the converse is not true: you cannot make glass out of either pottery or porcelain. The most probable explanation of the passage seems to be that Ho Ch´ou (who was apparently not a potter) experimented at some pottery with the materials used in glazing the green ware, and found that he could make a very good glass (_liu-li_) with the potter's green glaze, and perhaps other ingredients, a result which is in no way surprising, seeing that the softer ceramic glazes have a very close affinity to glass. But no further inferences can be drawn from this passage, and it is not even clear that Ho Ch´ou made a ceramic ware at all. All we are told is that he made _liu-li_. I have rather laboured this negative point, because Professor Zimmermann has published a declaration of belief that Ho Ch´ou was the discoverer of porcelain.[307] Apart from the obvious criticism which the writer himself anticipates, that such an epoch-making discovery would hardly have escaped the notice of Ho Ch´ou's biographer, Professor Zimmermann opens his case with a fundamental error, for which he has to thank Dr. Bushell. It is true that he only names Julien as the source of his information, but his version of the story of Ho Ch´ou is taken verbatim from Bushell's _Oriental Ceramic Art_,[308] where the crucial passage is unfortunately rendered "but he (Ho Ch´ou) succeeded in making vessels of green porcelain which could not be distinguished from true glass." This mistranslation puts an entirely different complexion on the passage, and goes a long way to justify Professor Zimmermann's inferences that Ch´ou made a glassy ware of the nature of porcelain. It is an instructive instance of the pitfalls which beset the student of Chinese subjects, especially when he has to rely on other people's translations.

Strange to say, a similar mistranslation occurs in Dr. Hirth's short but excellent treatise on _Ancient Chinese Porcelain_,[309] in a passage which is nevertheless of great importance to our quest. It has been the custom with Chinese compilers of reference works to incorporate the material of previous editions, adding their own commentaries and any further information which happened to have reached them, and to this we are indebted for the preservation of many passages from ancient writers which would otherwise be extremely difficult of access. Thus Hirth found embalmed in the Sung Pharmacopœia two early references to the material _pai o_ [Chinese] which he shows to be without doubt the kaolinic earth used in the manufacture of porcelain, and which, like many other strange materials, entered into Chinese medicinal prescriptions. The first mention of this substance is taken from the writings of T´ao Yin-chü, who died in 536 A.D., to the effect that the _pai o_, besides being used in medicines, was employed at that time for painting pictures; and Hirth argues that so celebrated a writer on scientific subjects as T´ao Yin-chü could not have failed to note it if the _pai o_ had been in general use for ceramic purposes as well. This is followed by a quotation from the T´ang Pharmacopœia (compiled about 650 A.D.): "It (_pai o_) is now used for painter's work, and rarely enters into medicinal prescriptions; during recent generations it has been prepared from white ware[310] (_tz´ŭ_)." By rendering the last sentence "during recent generations it has been used to make white porcelain," Hirth invested the passage with a greater interest than it actually possesses. But even when stripped of this fictitious importance, it constitutes the first literary evidence we have of the use of kaolin by Chinese potters. This is followed by another quotation from the T´ang Pharmacopœia recommending for medicinal purposes a powder prepared from the white ware of Ting Chou.[311]

Whether we are to understand that the Chinese pharmacist ground up broken pieces of Ting ware or merely made use of the refined and purified clay obtained at the potteries, matters little. Neither proceeding would be without parallel in Europe in far later times than the T´ang period. But the specific reference to white Ting ware at this early date is most interesting in view of the fact that Ting Chou was celebrated in the Sung dynasty for a white ware which is undoubtedly a kind of porcelain.

The presence of a kaolin-like material in a dark-coloured ware, probably of the third century, which was disclosed by the analysis made by Mr. Nicholls in Chicago, has already been recorded (p. 15). We have no means of ascertaining what length of time elapsed before a white material of this nature was evolved, but it was clearly in existence in the beginning of the sixth century. Possibly it was not porcelain according to the strict European definition, but there is every reason to suppose that it was a hard white ware, such as the Chinese would not hesitate to include in their porcelain category. Such a ware appears on some of the funeral vases which may safely be referred to the early T´ang period (see p. 26), and in default of other evidence I think we can say that porcelain in the Chinese sense already existed at the end of the Sui dynasty.[312]

Though this period happens to coincide with the lifetime of Ho Ch´ou, neither his name nor any other has been associated with the event by the Chinese, and it is highly probable that porcelain only came into being by a process of evolution from pottery and stoneware, the critical moment arriving with the discovery of deposits of kaolinic earth. As a mere speculation, I would suggest that the deposits were those at Han tan, the modern Tz´ŭ Chou, which supplied material for the Ting Chou potters.[313] It is, at any rate, significant that the new name, which we are led to suppose was derived from the _tz´ŭ_ stone,[314] was given to that place in the Sui dynasty.

Numerous literary references from this time onwards have already been quoted which are highly suggestive of porcelain. The "false jade vessels" of T´ao Yü in the early years of the seventh century; the eighth-century tea bowls of Yu Chou and Hsing Chou which were compared respectively to jade and ice, to silver and snow, the former being green and the latter white. The twelve cups used for musical chimes by Kuo Tao-yüan; the white bowls immortalised by the poet Tu "of ware (_tz´ŭ_) baked at Ta-yi, light but strong, which gives out a note like jade when struck."

The quality of translucency which in Europe[315] is regarded as distinctive of porcelain is never emphasised in Chinese descriptions. I can find no mention of it in any of the earlier writings,[316] and the first unmistakable literary evidence of its existence[317] comes from a foreign source. The Arab traveller, Soleyman, who describes his experiences in China in the ninth century, states that "they had a fine clay (_ghādar_) from which bowls were made, and in the transparency of the vessels the light of the water was visible; and they were (made of) fine clay."[318] This statement practically proves the existence of translucent porcelain in the T´ang dynasty, and we confidently await the arrival of specimens from Chinese excavations. Some of the export porcelains of the time have been actually unearthed at Samarra on the Euphrates by Professor Sarre,[319] of the Kaiser Friederik Museum, Berlin, and they include (1) bowls of gummy white porcelain with unglazed gritty base; (2) greenish white ware; (3) yellowish white with small crackle; and (4) a pure white porcelain with relief designs, such as birds and fishes, under the glaze. Other pottery found on that site included mottled ware of typical T´ang type, and a creamy white of the Tz´ŭ Chou type with brown spots.[320]

From considerations of form and the general character of the ware, I am inclined to regard three specimens in Plates 44 and 45 as belonging to the T´ang period. Fig. 1 of Plate 44 has a thin ivory white glaze running in gummy drops and clouded with pinkish buff staining outside and with a reddish discoloration within. Fig. 2 of the same Plate is remarkable in many ways. It is so thin as to seem to consist of little else but glaze, and is consequently almost as translucent as glass. The colour of the glaze is pearly white powdered with tiny specks, and the crackle is clearly marked. The base is flat and discloses a dry white body of fine grain. But its most conspicuous feature is the arresting beauty of its outline, which recalls some choice specimen of Græco-Roman glass, and displays a classic feeling frequently observed in T´ang pottery and in the Corean wares which owe so much to T´ang models.

Fig. 2, Plate 45, shows the celebrated phœnix ewer belonging to Mr. Eumorfopoulos which has proved so difficult to classify.[321] It is a white porcellanous ware translucent in the thinner parts, and the glaze is of light greenish grey with a tendency to blue in places. The form and ornament show strong analogies with specimens of T´ang pottery. The neck, for instance, may be compared with Fig. 2 of Plate 14; the phœnix head and the foliate mouth with Fig. 1 of Plate 9, and the carved ornament on the body with Fig. 3 of Plate 14.

Among the Sung wares many of the white Ting specimens are found to be translucent in their thinner parts, and these may be fairly regarded as porcelain proper. A considerable number of other white porcelains have come over of late under the description of Sung wares, and many of them are certainly early enough in form and style to belong to that period. They are true hard porcelain, translucent, and of a creamy white colour. Being for the most part without decoration, they can only be judged by their forms, and in view of the conservative habits of the Chinese, it would be rash to assert too emphatically their Sung origin. Yüan and even early Ming dates are suggested by the more cautious critics; but the possibility of a Sung origin having been established, I am inclined to give the evidence of form its full weight.

There are, besides these, other well-defined types of translucent porcelain which may confidently be attributed to a period as early as the Sung, but here the possibility--nay, probability--of a Corean origin has to be considered. It is certain that many of them have been found in Corean tombs; the provenance of the rest is doubtful. One type is a delicious smooth white porcelain with glaze of faintly bluish tinge, highly translucent, and worked very thin at the edges. The base of the vessels (usually small shallow bowls or saucers) is unglazed, and shows a soft-looking sugary body of close texture, rather earthy than glassy, and slightly browned by the fire. They have, in fact, almost the appearance of a "soft-paste" porcelain like that of Chelsea. These are so different from any known Chinese type that I strongly incline to a Corean origin for them. Another type is of hard but translucent ware with glaze of distinctly bluish tinge. A bowl in the British Museum is a good example of this. Of the usual conical form, it has a plain outside, and the inside is decorated with an incised design of not very clear meaning, but apparently a close foliage ground with highly formalised figures of boys. If this interpretation is correct, it is a conventional rendering of the well-known pattern of boys in foliage, Chinese in origin, but frequently used by the Corean potters. The design ends in Corean fashion, about an inch below the rim, leaving a plain band above it. The glaze is a faint bluish colour all over, and is powdered with specks, a fault in the firing; the base is almost entirely unglazed, and the biscuit, where exposed, has turned reddish brown. The style of this piece is strongly Corean. The same peculiarities in the base are shared by another type of small bowl, usually decorated inside with a sketchy design in combed lines. Some of these are creamy white; others are bluish white with a decided blue tinge in the well of the bowl where the glaze has formed thickly.[322] Another group, which is also said to be represented among the Corean tomb wares, is practically indistinguishable from the creamy white Ting Chou porcelain.

[Illustration:

Fig. 1

Fig. 2

Plate 44.--Early Translucent Porcelain, probably T´ang dynasty.

Fig. 1.--Cinquefoil Cup with ivory glaze clouded with pinkish buff stains. Diameter 3¾ inches. _Breuer Collection._

Fig. 2.--Vase of white, soft-looking ware, very thin and translucent with pearly white, crackled glaze powdered with brown specks. Height 3⅛ inches. _Peters Collection._]

[Illustration:

Fig. 1

Fig. 2

Plate 45.--T´ang and Sung Wares.

Fig. 1.--Square Vase with engraved lotus scrolls and formal borders. _T´u-ting_ ware, Sung dynasty. Height 20 inches. _Peters Collection._

Fig. 2.--Ewer with phœnix head, slightly translucent porcelain with light greenish grey glaze with tinges of blue in the thicker parts; carved designs. Probably T´ang dynasty. Height 15¼ inches. _Eumorfopoulos Collection._]

The whole question of these interesting porcelains is complicated by the fact that the Coreans were admittedly indebted to the Ting Chou potters for many of their designs;[323] and by the fact that while close intercourse between China and Corea existed the Coreans may well have imported Chinese wares and deposited some of them in the tombs. An authentic find of these porcelains in a Chinese tomb would give important evidence on this point, but so far there is no evidence of their being found in China beyond the statement of traders, and it is quite certain that they have been found in Corean tombs. It may be added that the Japanese class them as _hakugorai_ or white Corean ware, and stoutly support their Corean origin.

FOOTNOTES:

[Footnote 292: Op. cit., pp. 17-20.]

[Footnote 293: _Chinese Art_, vol. ii., p. 17, and the _Catalogue of the Morgan Collection_ (1907), p. xlviii.]

[Footnote 294: [Chinese] _wa ch'i yeh_. The _Shuo Wên_ was compiled by Hsü Shên and published first in 120 A.D. The word _tz'ŭ_ [Chinese] is compounded of the radical _wa_ [Chinese] (a tile, earthenware), and the phonetic _tz'ŭ_ [Chinese] (second, inferior), and carries no inherent suggestion of porcelain. If connoting a new material, it may be a name applied specially to glazed pottery which seems to date from the Han period, or even to stoneware as opposed to soft earthenware or brick.]

[Footnote 295: Thus the author of the _T'ang shih ssŭ k'ao_ (quoted in the _T'ao lu_, bk. viii., fol. 9 verso): "The characters [Chinese] and [Chinese] are not interchangeable. The latter is a hard and fine kind of _t'ao_. The material from which it is made is clay. The former [Chinese], on the other hand, is the name of a real stone which comes from the ancient Han-tan, which is the modern Tz'ŭ Chou. This department has potteries in which they use the _tz'ŭ_ stone for the body of the ware. Hence the name _Tz'ŭ ch'i_ (Tz'ŭ wares), not that the ware from the potteries of this place is all porcelain. I hear that at Ching-tê Chên the common usage is to employ the character [Chinese] for porcelain in writing and speaking. I have consulted friends whom I meet, and many use the two terms interchangeably. Truly this is altogether ridiculous. Tz'ŭ Chou is still making pottery at the present day." For the Tz'ŭ Chou pottery, see ch. viii.]

[Footnote 296: _Yeh chih_, quoted in the _T'ao lu_, bk. ix., fol. 13 recto.]

[Footnote 297: _T'ao shuo_, translated by Bushell, op. cit., p. 95.]

[Footnote 298: Bushell, _Chinese Art_, vol. ii., p. 18.]

[Footnote 299: See _T'ao shuo_ (Bushell, op. cit., pp. 97 and 99). See also bk. ix., fol. 1 verso, where the passage from the Annals of the Sui Dynasty is quoted.]

[Footnote 300: [Chinese] See _T'ao shuo_, bk. iv., fol. 17 recto and verso.]

[Footnote 301: The first two are apparently unidentified, but Jih-nan is Cochin China, whither, no doubt, the substance came as an article of trade.]

[Footnote 302: Early writers refer to it as _pi liu li_, which is a transcription of the Sanskrit _Vaidurya_, a stone supposed to be of the beryl type, but the identification is a matter of dispute. See Laufer, _Jade_, p. 111, footnote.]

[Footnote 303: A seventh-century writer.]

[Footnote 304: The _Ta Yüeh-chih_ have been identified with the Massagetæ, who in the fifth century were in possession of Afghanistan. See Bushell, _T'ao shuo_, op. cit., p. 100.]

[Footnote 305: The substance is discussed at length in connection with _pi-liu-li_ by Laufer (_Jade_, pp. 109-112), but this author seems very loath to admit the meaning glass for _liu-li_, though he allows that it is a common term for ceramic glaze. But the passage quoted above from the _T'ao shuo_ can hardly be explained in any other way than in reference to a kind of glass.]

[Footnote 306: The exact words of the text are [Chinese] (_Ch'ou i lü tz'ŭ wei chih yü chên wu i_). "Ch'ou took green ware and made it (_liu-li_) not different from the real."]

[Footnote 307: _Orientalisches Archiv_, Bd. ii., 1911, and _Chinesisches Porzellan_, p. 24.]

[Footnote 308: Op. cit., p. 20. Dr. Bushell, in his translation of the _T´ao shuo_, has given the more correct rendering, "Ch´ou made _some_ (i.e. _liu-li_) of green porcelain."]

[Footnote 309: Op. cit., pp. 3 and 4.]

[Footnote 310: [Chinese] _Chin tai i pai tz´ŭ wei chih_ = "in recent generations with white ware they make it."]

[Footnote 311: [Chinese] _Ting chou pai tz´ŭ._ For the Ting Chou ware, see ch. vii.]

[Footnote 312: Père d'Entrecolles (in his letter dated from Ching-tê Chên in 1712) makes the statement that the district of Ching-tê Chên sent regular supplies of its ware, which he terms porcelain, to the Emperor from the second year of the reign of _Tam ou te_ (sic). Though he gives the date as 422, it is clear that he really refers to the first Emperor, Wu Tê, of the T´ang dynasty (618-627 A.D.). It is not clear how he arrived at the conclusion that the ware in question was porcelain, and as he refers to the Annals of Fou-liang as his authority, we may assume that the Chinese phrase contained the inconclusive term _tz´ŭ_. or _t´ao_. He adds that "nothing is said as to the inventor, nor to what experiments or accident the invention was due."]

[Footnote 313: See p. 101.]

[Footnote 314: See p. 142.]

[Footnote 315: The European definition of porcelain may be stated thus: "Porcelain comprises all varieties of pottery which are made translucent by adding to the clay substances some natural or artificial fluxing material." In China the usual constituents are _kaolin_, which forms the clay substance, and _petuntse_ (china stone), which is the natural fluxing material. I should add that it is doubtful whether we are strictly justified in using the word _kaolin_ as a general name for porcelain earth (_o t´u_); but the term has been consecrated by usage, and has practically passed into our language in this sense. A slight translucency is observable near the rim on a white T´ang cup in the Eumorfopoulos Collection. The body of this piece is a soft white material, and the translucency is caused by a mingling of the glaze with the body where it is very thin, and it may be compared with the translucency of the Persian "gombroon" ware. But neither of these wares can be ranked as porcelain proper.]

[Footnote 316: It is, however, mentioned in connection with some of the Sung wares (the Kuan, for example), but only in relation to the glaze.]

[Footnote 317: It is true that Bushell, in his translation of the _T´ao shuo_ (op. cit., p. 102) implies this quality in a "brown ware (_tz´ŭ_) bowl" sent as tribute by the P´o-hai in 841 A.D. which is described as "translucent both inside and outside, of a pure brown colour, half an inch thick but as light as swan's down." The words of the text [Chinese] _nei wai t´ung jung_ ("inside and out throughout lustrous") are in themselves capable of suggesting translucence, but the remaining features--the brown glaze and the great thickness--are sufficient to preclude the idea of a translucent ware; and I imagine that the quality of lustre or translucency here applies only to the glaze. The P´o-hai appear to have been a subject state of Corea.]

[Footnote 318: I am indebted for this literal translation of the much-quoted passage to Mr. Edwards, of the Oriental MSS. Department of the British Museum. It has been more freely rendered by M. Reinaud, _Relation des voyages faits par les Arabes_, etc., Paris, 1845, p. 34.]

[Footnote 319: See F. Sarre, "Kleinfunde von Samarra und ihre Ergebnisse," in _Islam_, July, 1914.]

[Footnote 320: Fragments of white porcelain with carved designs were found in some of the sites excavated by Sir Aurel Stein in Turfan, and there are fragments similar to the Samarra finds obtained from ancient sites in the Persian Gulf and now in the British Museum. But the evidence of these pieces is not conclusive, for the sites were inhabited for many centuries. That of Samarra, on the other hand, is most important, for the city was only of a mushroom growth, which began and ended in the ninth century. See also p. 134.]

[Footnote 321: See _Cat. B.F.A._, 1910, A 43.]

[Footnote 322: See _Cat. B.F.A._, 1910, F 9 and 14.]

[Footnote 323: See passage from Hsü Ch´ing's notes, p. 39.]