is standard, but were found to be almost exactly in
accord with an instrument of fixed pitch, which in turn was found to
be approximately eleven beats at variance with the pitch-pipe on c2.

Assuming that the recording and transcribing speeds of the machines
were the same, this would place the original singing almost exactly in
accord with the old "philosophical standard of pitch" which places c2
at 512 double vibrations per second. Though the singing was not always
in perfect accord with the notes set down in transcriptions, with the
exception of those very marked departures especially indicated in the
music, the variations were so slight that, so far as true intonation
goes, the performances were fully up to the standard of those of the
average natural singer.

Special ear tubes were used while transcribing the records, and resort
made to a special device wherewith any order of whole, or even part
measures could be consecutively played. Thus it was possible to
closely compare parts which were similar in either words or music.

In some of the records two or more voices can be distinguished singing
in unison. Such unisons are shown in the transcription by single
notes. No attempt has been made to indicate the several voices. But
when such single notes are shown accompanied by the word "solo,"
it is to be understood that all of the performers have dropped out
but one, probably the leader. When the voices split up into parts,
it is so notated in the music.

Primitive people display more or less timidity in giving their
songs for scientific purposes. Such timidity is especially apt
to be manifested in their attacks. In the _Da-eng_, Girls' Part
(Record J), the delayed attack at the beginning of each new verse
is very marked. The delay varies considerably from verse to verse,
as indicated by the number of beats rest shown at the ends of the
lines. Similar pauses are found in the Boys' Part of the same ceremony
(see Record A). These beats rest or pauses are not to be taken as
part of the legitimate rhythm, for it is more than likely that if the
singers were giving their songs in their regular ceremonial and the
performers unconscious of observation, these pauses would not occur.

In transcribing those songs which have several verses on the record,
the notation has been so arranged on the page that the measures line
up vertically, making comparison easy between corresponding measures
of the different verses.

To indica