participle is noticeable. The construction aliquid locare faciendum also shows this union, and very likely formed one of the steps towards the use of the gerund as a participle of necessity.

To conclude, it seems to me that a theory which agrees with all the facts in Latin and is not contradicted by comparative grammar must be the right one. It therefore seems certain that the gerundive with its family bundus and cundus has been developed in the same manner as the other series, namely, by successive further formations, resulting in a verbal adjective (active or passive), and that this adjective has been attached to the verb, first as a present passive participle which the Latin had lost, then becoming a future passive participle (?), and finally a participle of necessity, as in its use as nominative and accusative. That, further, the gerund (as is generally recognized) is nothing more than the impersonal of the gerundive taking a case according to the other uses of Tonтéov and agitandum.

THE

THE MOUTH-PIECE OF THE Avλós.

BY ALBERT A. HOWARD.

HE description given by Theophrastus, Hist. Plant. iv. 11, of the way in which the mouth-piece of the avλós was made, is apparently much more detailed and more complete than has hitherto been supposed to be the case.

2

The mouth-piece there described is, beyond any reasonable doubt, a double reed of the kind used in the modern bassoon. This is shown by a variety of evidence,1 and particularly by the technical name Çevyos2 (cf. §§ 4 and 6), which must mean a mouth-piece made of a pair of similar parts. That this pair was a pair of reeds (γλῶτται) is shown by the words τὸ στόμα τῶν γλωττῶν 3 at the end of § 4, for if these reeds belonged to two separate pipes the plural τὰ στόματα would have to be used.

The further evidence in Theophrastus would seem to indicate that the mouth-piece was, in his time, made in exactly the same way in which, after a lapse of more than two thousand years, the mouthpiece of the bassoon is now made. The process to-day is as follows. From a piece of cane, twice the length of the desired mouth-piece, a strip of the requisite width is split and the interior pulpy surface is

1 Cf. The Avλós or Tibia, Harv. Stud. iv. pp. 23-25.

2 The definition of ɲeûyos given in the lexicon of Liddell and Scott is certainly incorrect in so far as it relates to the avλós. In the singular the word implies a pair, and it would be absurd to use the plural ševyn to denote a pair of pipes. Furthermore, in the passage cited in support of this definition, Theophrastus is speaking, not of the instruments themselves, but of the mouth-pieces, as is clear from the context and particularly from § 6, where the cane has been cut into strips about six inches long (much too short for any instrument). In the other writers who use this word the reference is always to the mouth-piece and never to the instrument. The plural evyn must therefore mean mouth-pieces, ordinarily a pair of them.

3 Cf. von Jan's emendation of Aristotle, de audib. p. 801 b 33, тà YÀρ ĕXOVTA TŵV Jèvyéwv (MS. deutépwv) Tàs yλwoσas #λayías, etc., Berl. Phil. Woch., 1894, col. 209.

gouged out. The strip is next scraped in the middle of its outer surface until it will bend easily lengthwise; the scraped part is well moistened, and the ends of the strip are bent round until they touch each other. The ends thus brought together are wound with strong thread and curved by pressure into a cylindrical tube which can be attached to the instrument. The two flattened surfaces caused by bending the strip of cane are then scraped very thin and the mouthpiece is finished by cutting the strip at the bend.

Theophrastus tells us in § 6 that the strip of cane, from which in his time the mouth-piece was made, was at least two palms, 14.8 cm. long, a length which is nearly one fourth that of the longest pipe found at Pompeii, and nearly one half that of either of the Elgin pipes in the British Museum. It will be readily seen that this length is out of all proportion to the length of the pipes themselves.

The representations in works of art invariably show a short mouthpiece, and my own experiments in determining the scales of the ancient instruments which have been preserved led me to choose a mouth-piece not longer in any case than 8 cm. Two mouth-pieces of Egyptian pipes, one of which is certainly, the other probably, of the double-reed type, found in the ruins of Panopolis and described1 by Victor Loret, are respectively 7.6 cm. and 8 cm. in length. The mouth-piece of the modern bassoon is about 7 cm. in length.

This evidence leads naturally to the conclusion that the finished mouth-piece of the avλós was only about half as long as the strip of cane described by Theophrastus. Furthermore, both reeds of the (cuyos were made from the same joint or strip of cane, as is shown by § 7, συμφωνεῖν δὲ τὰς γλώττας τὰς ἐκ τοῦ αὐτοῦ μεσογονατίου τὰς δὲ aλas où σvμowveîv; while the words which follow, describing the

1 One of these mouth-pieces is described in the Journal Asiatique for 1889, p. 213, the other in a paper read before the Société d'anthropologie de Lyon, June 3, 1893. The one described in the second paper is of the double-reed type, and though somewhat longer, closely resembles the double reed of the modern bassoon. It is of course possible that the mouth-pieces of Greek and Egyptian pipes were not made in the same way, but it is hardly probable since the instruments themselves are very like each other.

2 I interpret these words to mean that reeds from the same joint of cane, if made into a mouth-piece, will vibrate in unison, but that others will not. The

final operation in the making of the mouth-piece, seem to imply that the strip of cane has been doubled upon itself : τμηθέντος δὲ δίχα τοῦ μεσογονατίου τὸ στόμα τῆς γλώττης ἑκατέρας γίνεσθαι κατὰ τὴν τοῦ kaλáμοv Toμýv, for if the strip is cut at the point where it was bent, the mouth of each reed is at the point where the cut was made, agreeing exactly with the statement of Theophrastus.

The only apparent objection to this explanation is caused by the words καὶ τὴν μὲν (γλώττην) πρὸς τῇ ῥίζῃ ἀριστερὰν εἶναι τὴν δὲ πρὸς τοὺς βλαστοὺς δεξιάν, which follow immediately the words τὰς δὲ ἄλλας οὐ συμφωνείν, and have always been interpreted as meaning respectively the reed of the left and the reed of the right avλós. This interpretation is perhaps due to a passage in Pliny, N. H. xvi. 172, sed tum ex sua quemque tantum harundine congruere persuasum erat et eam quae radicem antecesserat laevae tibiae convenire, quae cacumen dexterae, which seems to be an attempt to give the substance of the statement in Theophrastus.1 The words åpioTepáv and deέiáv are, however, feminine, and must refer to yλúrтηv, not to avλóv. It is hardly conceivable that Theophrastus should have been so careless in his use of language as to speak of right and left reeds when he meant the reed of the right pipe and the reed of the left pipe. It is far more likely either that Pliny misunderstood the passage of Theophrastus in this point, as he certainly did in others, or that he copied it, without verification, from carelessly written notes and excerpts such as are described in the letters of the younger Pliny, iii.

5. II.

The whole difficulty disappears at once, however, if we suppose that Theophrastus had before him, while writing, an illustration either of the plant itself or of one of the joints suitable for making a mouth

reeds must be those of a single feûyos and not those of a pair of jeúyn, for on any other supposition the statement of Theophrastus is obviously untrue, as is shown by every orchestral performance in which two oboes or two bassoons are employed, since the mouth-pieces of the two instruments are probably never made even from the same cane. The two reeds of a single mouth-piece are, however, always made from the same joint of cane, not for the reason given by Theophrastus, but because it is easier so to make them.

1 Owing to the fact that this passage is quoted by Pliny, we are shut off from the regular refuge in cases of difficulty, explaining the objectionable words in the Greek as a gloss.