382 NOTES

dots when uncertain (:). The latter kind are typical of the oldest ballad
tunes in the east and west. Hustvedt indicates sharps and flats and
naturals by the usual musical signs. These are not on most typewriters,
and I find that some confusion may arise through the fact that on staff
notation and with Mr. Hustvedt the accidental is entered before the note,
but in usual transcription it comes after. I think accents may serve. The
acute (') easily represents a sharp; the grave (x) a flat; and the tittle (~) a
natural, since it was originally a superimposed *ns.

By these means most simple melodies may be printed or typed within
the dimension of a square inch; an enormous saving on the space occupied
by staff notation.

The tune so transcribed is characterized by a melodic contour. This
contour can be set down in a quasi-algebraical way by counting each
rise and fall in semitones, ignoring identical notes. By using roman and
italic (or underlined) numbers one can get symbols for twenty rising
semitones: i-o j-o, where V is ten and £ o' is twenty. The alphabet
gives symbols for twenty-four falling semitones, omitting I and O
as liable to confusion with numbers. Actually one only needs about
twelve.

These formulas may be used for the whole tune or any part. For pur-
poses of comparison it is enough to give the curve of the first phrase,
corresponding to the first line of quatrains or, sometimes, the first hemi-
stich of long divided lines. In a general catalogue Mr. Hustvedt recom-
mends entering in numerical order first, and then in alphabetical order;
so tunes in n . . ., 12 . . ., 13 . . ., &c., then Ai . . ., Az . . ., and later
AA . . ., AB . . . There would have to be cross-indexing, however, for it
happens that the first interval is often the most capricious. Singers make
different jumps to reach the same continuing pitch. In this way tunes are
sometimes identical, except for the first interval. I am not quite certain
whether all tunes which have identical melodic contours under Hustvedt's
transcription are in fact identical. The details can be ascertained by con-
sulting the full transcription.

Personally, I find it convenient to indicate in transcription by a blank
where each line ends. To make the above clear, I give a transcription and
characterization of God Save the King ( = My Country 'tis of Thee) as it
would appear on a card-index.

God Save the King z

Church Hymnary 631

I sharp 44 Uz

gga/f (.g)a/

bbc/b(.a)g/

agf/g-/

ddd/d(.c)bl

ccc/c(.b)&/