But what weight can be given to Aristides as an authority on the music of the time of Plato? The answer to this question depends upon several considerations.

1. The date of Aristides is unknown. He is certainly later than Cicero, since he quotes the De Republica (p. 70 Meib.). From the circumstance that he makes no reference to the musical innovations of Ptolemy it has been supposed that he was earlier than that writer. But, as Aristides usually confines himself to the theory of Aristoxenus and his school, the argument from silence is not of much value. On the other hand he gives a scheme of notation containing two characters, sharp and sharp, which extend the scale two successive semi-tones beyond the lowest point of the notation given by Alypius [43]. For this reason it is probable that Aristides is one of the latest of the writers on ancient music.

2. The manner in which Aristides introduces his information about the Platonic Modes is highly suspicious. He has been describing the various divisions of the tetrachord according to the theory of Aristoxenus, and adds that there were anciently other divisions in use. So far Aristides is doubtless right, since Aristoxenus himself says that the divisions of the tetrachord are theoretically infinite in number (p. 26 Meib.),—that it is possible, for example, to combine the ParhypatÊ of the Soft Chromatic with the Lichanos of the Diatonic (p. 52 Meib.). But all this concerns the genus of the scale, and has nothing to do with the species of the Octave, with which Aristides proceeds to connect it. It follows either that there is some confusion in the text, or that Aristides was compiling from sources which he did not understand.

3. The Platonic Modes were a subject of interest to the early musical writers, and were discussed by Aristoxenus himself (Plut. de Mus. c. 17). If Aristoxenus had had access to such an account as we have in Aristides, we must have found some trace of it, either in the extant Harmonics or in the quotations of Plutarch and other compilers.

4. Of the four scales which extend to the compass of an octave, only one, viz. the Dorian, conforms to the rules which are said by Aristoxenus to have prevailed in early Greek music. The Phrygian divides the Fourth a-d into four intervals instead of three, by the sequence a b b* c d. As has been observed, it is neither the Enharmonic Phrygian species (c e e* f a b b* c), nor the Diatonic d-d, but a mixture of the two. Similarly the Mixo-lydian divides the Fourth b-e into four intervals (b b* c d e), by introducing the purely Diatonic note d. The Lydian is certainly the Lydian Enharmonic species of the pseudo-Euclid; but we can hardly suppose that it existed in practical music. Aristoxenus lays it down emphatically that a quarter-tone is always followed by another: and we cannot imagine a scale in which the highest and lowest notes are in no harmonic relation to the rest.

5. Two of the scales are incomplete, viz. the Ionian, which has six notes and the compass of a Seventh, and the Syntono-lydian, which consists of five notes, with the compass of a Minor Sixth. We naturally look for parallels among the defective scales noticed in the Problems and in Plutarch's dialogues. But we find little that even illustrates the modes of Aristides. The scales noticed in the Problems (xix. 7, 32, 47) are hepta-chord, and generally of the compass of an octave. In one passage of Plutarch (De Mus. c. 11) there is a description—quoted from Aristoxenus—of an older kind of Enharmonic, in which the semitones had not yet been divided into quarter-tones. In another chapter (c. 19) he speaks of the omission of the TritÊ and also of the NÊtÊ as characteristic of a form of music called the spondeiakos tropos. It may be said that in the Ionian and Syntono-lydian of Aristides the Enharmonic TritÊ (b*) and the NÊtÊ (e) are wanting. But the ParamesÊ (b) is also wanting in both these modes. And the Ionian is open to the observation already made with regard to the Phrygian, viz. that the two highest notes (c d) involve a mixture of Diatonic with Enharmonic scale. We may add that Plutarch (who evidently wrote with Aristoxenus before him) gives no hint that the omission of these notes was characteristic of any particular modes.

6. It is impossible to decide the question of the modes of Aristides without some reference to another statement of the same author. In the chapter which treats of Intervals (pp. 13-15 Meib.) he gives the ancient division of two octaves, the first into dieses or quarter-tones, the second into semitones. The former of these (hÊ para tois archaiois kata dieseis harmonia) is as follows:

Two Octaves

After every allowance has been made for the probability that these signs or some of them have reached us in a corrupt form, it is impossible to reduce them to the ordinary notation, as Meibomius sought to do. The scholar who first published them as they stand in the MSS. (F. L. Perne, see Bellermann, Tonleitern, p. 62) regarded them as a relic of a much older system of notation. This is in accordance with the language of Aristides, and indeed is the only view consistent with a belief in their genuineness. They are too like the ordinary notation to be quite independent, and cannot have been put forward as an improvement upon it. Are they, then, earlier? Bellermann has called our attention to a peculiarity which seems fatal to any such claim. They consist, like the ordinary signs, of two sets, one written above the other, and in every instance one of the pair is simply a reversed or inverted form of the other. With the ordinary signs this is not generally the case, since the two sets, the vocal and instrumental notes, are originally independent. But it is the case with the three lowest notes, viz. those which were added to the series at a later time. When these additional signs were invented the vocal and instrumental notes had come to be employed together. The inventor therefore devised a pair of signs in each case, and not unnaturally made them correspond in form. In the scale given by Aristides this correspondence runs through the whole series, which must therefore be of later date. But if this is so, the characters can hardly represent a genuine system of notation. In other words, Aristides must have been imposed upon by a species of forgery.

7. Does the fragment of the Orestes tell for or against the Modes described by Aristides?

The scale which is formed by the notes of the fragment agrees, so far as it extends, with two of the scales now in question, viz. the Phrygian and the Dorian. Taking the view of its tonality expressed in the last chapter (p. 93), we should describe it as the Dorian scale of Aristides with the two highest notes omitted. The omission, in so short a fragment, is of little weight; and the agreement in the use of an additional lower note (Hyper-hypatÊ) is certainly worth notice. On the other hand, the Dorian is precisely the mode, of those given in the list of Aristides, which least needs defence, as it is the most faithful copy of the Perfect System. Hence the fact that it is verified by an actual piece of music does not go far in support of the other scales in the same list.

If our suspicions are well-founded, it is evident that they seriously affect the genuineness of all the antiquarian learning which Aristides sets before his readers, and in particular of his account of the Platonic modes. I venture to think that they go far to deprive that account of the value which it has been supposed to have for the history of the earliest Greek music.

For the later period, however, to which Aristides himself belongs, these apocryphal scales are a document of some importance. The fact that they do not agree entirely with the species of the Octave as given by the pseudo-Euclid leads us to think that they may be influenced by scales used in actual music. This applies especially to the Phrygian, which (as has been shown) is really diatonic. The Ionian, again, is perhaps merely an imperfect form of the same scale, viz. the octave d-d with lower d omitted. And the Syntono-lydian may be the Lydian diatonic octave c-c with a similar omission of the lower c.