MUSICAL IXTERXUJDE 131

* Exactly the same ? You would not prefer something else ?'

'No, I want him to repeat it exactly as he has just sung it.'

The request was communicated to the Ancient. He nodded ; he
began again. And now—note this well!—there were very wide
variations in the performance. True, it was mostly in five-eight
time, but there were three bars of two-four interposed for no
conceivable reason. In addition, half-way through the phrase
there was an elaborate cadenza which had not even been suggested
on the first variation.

Once more I held up my hand ; once more there was silence.

*That was exactly the same as the first time ?? I inquired,

'Exactly the same. Did you not hear it ? *

Oh yes, I had heard it. And I had heard something else besides.
I had heard the secret of Indian music.

VI
The reader who is bored by aesthetic theory must start skipping
again> though I wish he would try tp grin and bear it, and read on.
There is no reason why the critic of the arts should not be as
exciting as the writer of detective stories. If you are on the track
of an artistic crime the clues are as subtle, and the villains as
contemptible—while the heroine is Beauty immaculate.
The bebt detective story that Edgar Allan Poe ever wrote was a
literary essay. It was called fc The Rationale of Verse'; it described
how he came to write 'The Raven'; and the manner in which he
tracked down that awesome bird to its ultimate immortal perch
was breathlessly exciting, even though his only instruments were
dactyls and spondees, rhymes, and refrains.
So let us consider our little discovery as a clue in a mystery. The
second version of the Ancient's song was quite different from the
first, yet he claimed that it was exactly the same. Where does the
discrepancy lead us ?
It leads us, as we suggested above, straight to the heart of the
secret. And that secret lies in the word improvisation. The Ancient
was not singing a definite role, he was ixot interpreting a theme