the shape of the sounding-board always remained the same; the outer side was a good deal raised, while the inner side on which the strings were attached was a level surface. Into this sounding-board two arms were fixed, which are almost always represented on Greek monuments as merely curved pieces of wood fastened on the inner side of the sounding-board; but the custom which in later times, especially in the Alexandrine and Roman periods, became very common, of not merely constructing these arms in the shape of horns, but even making them of real horns of chamois or gazelles, no doubt existed even in the ancient Greek period. At their upper ends the two arms, which might be called horns, were fastened together by a cross-piece, called the yoke, which was usually constructed of hard wood, and on to this the strings, constructed of sheep-guts, were stretched. Of these the lyre usually had seven, all of equal length, which was also the case in the cithara. These strings, as we can clearly see in the lyres of the above-mentioned bowl (Fig. 75), passed downwards over a bridge consisting of a piece of reed fixed on the flat covering of the sounding-board, and were then fastened singly, probably to a little square board, such as we see on the lyre hanging on the wall in Fig. 75. Probably this little board could be taken out, and thus, if a string were to break, the injury could be easily repaired. Occasionally the strings were merely tied to the yoke; but, as this primitive method would make it impossible to tune them, we must assume that there was usually some other contrivance, though neither writers nor monuments give us sufficient information about it. On the lyres in Fig. 75, and also in other pictures of stringed instruments, we perceive at the upper ends of the strings, longish rolls which in other places are shaped more like rings or discs, and are probably set at an angle to the stretched strings. An hypothesis has been set up by Von Jan, who infers, from ancient writers, after comparing similar contrivances in Nubian stringed instruments, that these rolls were constructed of thick skin or hide, taken from the backs of oxen or sheep; the strings were fastened into these adhesive covers and twisted along with them round the yoke of the lyre until they attained the right tune, and they were then fastened into their proper position by strongly pressing down these rolls of hide. Still, this rough mode of fastening which could only permit of very superficial tuning of the strings, does not appear very satisfactory; indeed, Von Jan himself calls attention to a far more artistic contrivance observed in some of the pictures which has not yet, however, been satisfactorily explained. There seems also to have been a third mode of fastening; sometimes the whole yoke was divided into as many little pulleys connected by pegs as there were strings, so that each string had, as it were, its own yoke, by the tightening of which it