the Chinese system is the most crudely backward and incapable of development of any of the great melodic systems. But at the same time it must not be ignored that notwithstanding such obstacles, and the fact that musicians are looked down upon as an inferior caste in China, the Chinese do manage to produce good and effective tunes; and it cannot be denied that the pure pentatonic system lends itself peculiarly to characteristic effects, and to the production of impressions which are more or less permanent. Its very restrictions give it an appearance of strangeness and definiteness which attract notice, and with some people liking.

Nations which have not been so tied and bound by ordinances and dogmatic regulations have managed to develop pentatonic systems to a much higher degree of artistic elasticity, and the result has naturally been in some cases to minimise the characteristic pentatonic effect. The Japanese were among the foremost to expand their system in every practicable way. They have nominally as complete a series of twelve semitones as European musicians, but, like all other cultivators of melodic music, they only use them to select from. Authorities may be confessed to differ, but their scale-system seems to be pentatonic in origin, like that of the Chinese; though, unlike them, they distribute their intervals so as to obtain twelve different modes of five notes each. For instance, one mode of five notes, called Hiradioschi, corresponds to C, D, E, G, Ab; another, Kumoi, to C, D, F, G, A; another, Iwato, to C, D, F, G, B+; from which it is to be observed that they fully appreciate the artistic value of semitones; which again distinguishes them from the Chinese, who rarely use such intervals. They are said to make use of the octave, the fifth, and the fourth in tuning, and to tune their thirds and sixths by guesswork, and not by any means scientifically. The thirds are said to be often more like the "neutral thirds" described in connection with Persian music, which are neither major nor minor, but between the two. A Japanese musician,

*

Mr. Pigott gives a note equivalent to E. + Mr. Pigott gives a note equivalent to Bb.

who seems fully competent to form an opinion, has expressed doubts as to whether their scale was true pentatonic or not. In face of the distinct grouping of five notes which is almost invariable, this view seems rather paradoxical; but the frequent occurrence of a fourth with a semitone above the lower note is so like the early tetrachord of the Greeks, with a sensitive downward-tending leading note (see p. 23), that the doubt cannot be said to be without some appearance of justification. The mode Kumoi, quoted above, would in that sense represent two tetrachords, C, D, F-G, A C, like those of Olympos, put one above another; and the effect of them may be gauged by the process suggested on p. 20.

There are two other important systems of melodic music which are most probably true pentatonic, but quite different from either Chinese or Japanese. The oldest of them is the Javese. In this case there is no possibility of unravelling the process of development of the scales; we can only take the results as examined by Mr. Ellis and Mr. Hipkins, whose methods seem thoroughly trustworthy, and gather what we can from the facts. The Javese have two plans of tuning, one called Gamelan Salendro, and the other Gamelan Pelog, which differ so much that they cannot be played together. In the Gamelan Salendro scale there are five notes, which are fairly equidistant from one another, and each of the intervals exceeds a whole major tone, such as C and D, by a considerable interval. To our European ideas such a scale seems almost inconceivable. To compare it with our major scale of C, the first degree would be from C to a note halfway between D and E, the next degree would be between E and F but nearer to F, the next would be a quarter of a tone higher than G, and the next about half-way between A and B, and the next move would be to the octave C above the starting-point. How such a scale could be tuned by ear almost passes comprehension, and implies a very remarkable artificial development of scale-sense in the musicians who use it. The Gamelan Pelog is a very different mode, and almost as singular. The first step would be from C to a note a little higher than E, the second to a note a little below F,

the third note would be just below G, the next a little below B, and the remaining step would reach the octave C. This is evidently a very elaborate artificial development of some simpler pentatonic formula that has long passed out of record. The Siamese system is almost as extraordinary. It is not now pentatonic, though supposed to be derived originally from the Javese system. The scale consists of seven notes, which should by rights be exactly equidistant from one another; that is, each step is a little less than a semitone and threequarters. So that they have neither a perfect fourth nor a true fifth in their system, and both their thirds and sixths are between major and minor; and not a single note between a starting note and its octave agrees with any of the notes of the European scale. The difficulty of ascertaining the scale used in practice lay in the fact that when the wooden harmonicon, which seemed the most trustworthy basis of analysis, was made out of tune, the Siamese set it right by putting pieces of wax on the bars, which easily dropped off. Their sense of the right relations of the notes of the scale is so highly developed that their musicians can tell by ear directly a note is not true to their singular theory. Moreover, with this scale they have developed a kind of musical art in the highest degree complicated and extensive.

This survey would not be complete without reference to the scale of the Scotch bagpipe. This, again, is a highly artificial product, and no historical materials seem available to help the unravelling of its development. Though often described as pentatonic, the scale comprises a whole diatonic series of notes, from which modes may be selected. These notes do not agree with our ordinary system, and their relations are merely traditional, as they are tuned empirically by ear. Taking A as a starting-point, the next note is a little below B; the next is not C, but almost a neutral third (p. 29) from A; the next very nearly a true fourth above A, that is, a little below our D; the next almost exactly a true fifth from A, that is, very near E; the next a neutral sixth from A (p. 29), between E and F; and the remaining note a shade below G. The type is more like the ancient Arabic than any

other, and not really the least like the Chinese, though the impression conveyed by the absence of the leading note sometimes misleads people into supposing they are akin. Whether it is really a pentatonic scale, as some have thought, is therefore extremely doubtful. Even if the modes were really of five notes, that is not a proof that its constitution is of the pentatonic order, as has been indicated in connection with the Indian and Japanese system; both the fifth and the fourth are very nearly true, and as it seems based on the old Arabic system, which was not pentatonic, the argument would tend to class it with the Indo-European and Persian seven-note systems.

The above summary is sufficient to show the marvellous variety of the scales developed by different nations for purely melodic purposes. The simple diatonic system of the Greeks, the subtly ingenious mathematical subdivisions of the Persians and Arabs, the excessive modal elaborations of the Hindus, the narrow and constricted stiffness of the Chinese, the ambiguous elasticity of the Japanese, and the truly marvellous artificiality of the Javese and Siamese systems, are all the products of human artistic ingenuity working instinctively for artistic ends. Similarity of racial type seems to have caused men to produce scales which are akin. They are all devised as means to ends, and when the mental characteristics and artistic feeling of the races who devised the scales have been similar the result has been so too. The seven-note systems are mostly characteristic of Caucasian races, and the five-note scales of the somewhat mixed but probably kindred races of Eastern Asia. And this does not so much indicate that they borrowed from each other as that the same types of mind working under artistic impulse produced similar results. One important defect they have in common. Though in most of them the relations of the notes are actually defined with the utmost clearness, in none have they arrived at the artistic completeness of maturity which is implied by classification. This remained to be done under the influence of harmony.

It is quite clear that the early Christians adopted the prin ciples and some of the formulas of melody of the ancient

Greek system-in the state to which it had arrived at about the beginning of our era-for as much music as their simpe ritual required. But none of it was written down, and in those centuries of general disorganisation in which the collapse of the Roman Empire was going on, the traditions became obscure and probably conflicting in different centres. To remedy this state of things efforts were made, especially by Ambrose, Bishop of Milan, and one of the many Popes named Gregory, to establish uniformity by restoring the system of the Greek modes and making the music they used conform to it. Knowledge of every kind was at that time at a very low ebb, and the authorities who moved in the matter had very limited and indefinite ideas of what Greek music had been. But between them they contrived to organise an intelligible arrangement of various modes, and it was of no great consequence that they got most of the names wrong. Ambrose authorised four modes, the (1) Dorian, (2) Phrygian, (3) Lydian, and (4) Mixolydian-corresponding more or less to the ancient Greek (1) Phrygian, (2) Doric, (3) Syntono-Lydian, and (4) Ionic. These were called the authentic modes. Gregory nominally added four more, which were not really new modes, but a shifting of the component notes of the modes of Ambrose; for as by Ambrose's regulations musicians were only allowed to use the scale of D between D and its octave, by Gregory's arrangement they might use the notes a, b, c below the lower D instead of in the higher part of the scale. And similarly with the other three. Gregory's group were called plagal modes. In later days four more modes were added: the mode beginning on C, and that beginning on A and their plagals; and two hypothetical modes which were not supposed to be used, namely, that beginning on B and its plagal. The total amounted therefore to fourteen modes, of which two were not actually used. It was very soon after this organisation of modes that attempts at harmony began to be made, either by doubling an ecclesiastical tune at another pitch, such as the fourth or the fifth, or by really trying to get two tunes to go together. The idea of harmony in the modern sense did not develop into clearness for centuries; but musicians got more