becomes a portion of himself. Before I even take up my violin I study a
piece thoroughly in score. I read and reread it until I am at home with
the composer's thought, and its musical balance and proportion. Then,
when I begin to play it, its salient points are already memorized, and
the practicing gives me a kind of photographic reflex of detail. After I
have not played a number for a long time it fades from my memory--like
an old negative--but I need only go over it once or twice to have a
clear mnemonic picture of it once more.

"Yes, I believe in transcriptions for the violin--with certain
provisos," said Mr. Elman, in reply to another question. "First of all
the music to be transcribed must lend itself naturally to the
instrument. Almost any really good melodic line, especially a
_cantilena_, will sound with a fitting harmonic development. Violinists
of former days like Spohr, Rode and Paganini were more intent on
composing music _out of the violin_! The modern idea lays stress first
of all on the _idea_ in music. In transcribing I try to forget I am a
violinist, in order to form a perfect picture of the musical idea--its
violinistic development must be a natural, subconscious working-out. If
you will look at some of my recent transcripts--the Albaniz _Tango_, the
negro melody _Deep River_ and Amani's fine _Orientale_--you will see
what I mean. They are conceived as pictures--I have not tried to analyze
too much--and while so conceiving them their free harmonic background
shapes itself for me without strain or effort.


A REMINISCENCE OF COLONNE

"Conductors with whom I have played? There are many: Hans Richter, who
was a master of the baton; Nikisch, one of the greatest in conducting