not be as neatly done in all respects as any other part. It may be that
the supposition is uppermost in the mind of the repairer that, like
the nut at the fingerboard, the pressure of the strings will retain
it in position. This is a mistake, there is a great pull forward,
especially if the wood is hard and dry. The material should be selected
for its solidity and hardness like that for the other nut. In olden
times, say those of the early Italians, this part, owing to the small
amount of strain in consequence of low pitch, low bridge and short neck,
seems to have been treated with almost indifference, a very slight
piece of ebony, cherry, pear, or other variety of hard wood found in
Italy, sufficing for the purpose (diagram 10). It was left level with
the surrounding soft wood, or nearly so; there was no occasion for
raising it at the time, as the tail-string projected from the
underneath of the tailpiece instead of that almost universally now
known as the secret tie (diagrams 11 and 12). This latter necessitates
the use of a higher and more substantial nut, otherwise the tailpiece
would be close down to, if not touching, the table, causing a rattling.
Further, in accordance with mechanical law, the strain or pull forward
increases with the height of the nut. It is therefore obvious, that
unless well fitted and held strongly, the nut will be liable to be
wrenched forward out of position. This is more frequent than would be
suspected, and is sometimes a secret source of damage or bad influence
leading to disaster in other parts of the instrument. The same
observations concerning the preliminaries apply to the fixing of this
as to the other nut. The modern arrangement of the part leaves but
little to be improved upon.

[Illustration: DIAGRAM 10. OLDEST FORM OF NUT.]

[Illustration: DIAGRAM 11.]

[Illustration: DIAGRAM 12.]

The length and thickness of the nut required having been determined
upon, we will suppose ourselves in the presence of an old worthy from
Cremona requiring a fresh attachment, the wood selected--Mauritius
ebony for preference--and the measurement as follows, 5/16 in length
and thickness according to the width of the border, as the nut looks
best when the inner edge runs in a continuous line with that of the
purfling (diagram 13). In highly finished work and when the end of the
violin has a perceptible curve instead of being nearly straight, the
nut should be made to follow the course of the purf