s companions. Among the earlier volumes
such use counts for little, owing to the large number of volumes averaged,
while it may and does make the figures for the later volumes irregular.
Thus, under History the high number in the twelfth column represents
one-twelfth volume of Froude, which was taken out three times, evidently
for separate reference, as the eleventh was withdrawn but once.
Furthermore, apart from this irregularity, the figures for the later
volumes are relatively large, for a work in many volumes is apt to be a
standard, and although its use falls rapidly from start to finish enough
readers persevere to the end to make the final averages compare unduly
well with the initial ones where the high use of the same work is averaged
in with smaller use of dozens of other first and second volumes. That the
falling off from beginning to end in such long works is much more striking
than would appear from the averages alone may be seen from the following
records of separate works in numerous volumes:

VOLUMES
I II III IV V VI VII VIII IX X
HISTORY

Grote, "Greece" 11 6 5 2 1 0 1 1 1 0
Bancroft, "United States" 22 10 6 8 10 8
Hume, "England" 24 7 5 2 1 1
Gibbon, "Rome" 38 12 7 3 4 6
Motley, "United Netherlands" 7 1 1 1
Prescott, "Ferdinand and
Isabella" 20 4 2
Carlyle, "French Revolution" 18 10 8
McCarthy, "Our Own Times" 27 8 11

BIOGRAPHY

Bourienne, "Memoirs of
Napoleon" 19 18 9 7
Longfellow's "Life" 6 4 2
Nicolay and Hay, "Lincoln" 6 3 3 2 2 2 2 1 1 2
Carlyle, "Frederick the
Great" 7 3 2 2 2

FICTION

Dumas, "Vicomte de
Bragelonne" 31 30 24 22 21 16
Dumas, "Monte Cristo" 27 17 18
Dickens, "Our Mutual Friend" 5 4 1 0
Stowe, "Uncle Tom's Cabin" 37 24

Of course, these could be multiplied indefinitely. They are sufficiently
interesting apart from all comment. One would hardly believe without
direct evidence that of thirty-one persons who began one of Dumas's
romances scarcely half would read it to the end, or that not one of five
persons who essayed Dickens's "Mutual Friend" would succeed in getting
thro