searching its surface for some further clue. At once he noticed the
arrangement of the concentric circles of letters which made up the Latin
prayer. The words were so written that each letter stood opposite a
pearl, and reading inward from each pearl, there was a row of letters
six deep reaching almost to the center of the box. Clearly here were six
different ciphers, that is, six circles of twenty-six letters each, any
one of which might constitute a working cipher. It was only necessary to
call the big pearl at the top "_A_," and here were six different letters
opposite it, any one of which, in a system of cipher writing, might be
used as the letter _A_.

Duvall, however, knew enough about ciphers to know that such an
arrangement constituted no cipher at all, in other words, that ciphers
so simple, so readily solved, as this, would never be employed in any
case where absolute secrecy was imperative. He felt that there was
something more to the matter than he had so far discovered.

Suddenly he saw that, just beyond each pearl, was engraved on the ivory
rim of the box a number--starting with the large pearl at the top as
number one, the circle of numbers ran around the edge of the box until
it returned to its starting point, at number twenty-six. In his efforts
to see these numbers, which were very small, he gripped the box tightly
in his hands to hold it the more steadily toward the rather dim light.
In doing so, he suddenly became aware of the fact that the rim or edge
of the box, containing the numbers and the circle of pearls, was
movable. It fitted so cunningly into the top of the box, that the joint
appeared not as a crack or perceptible space, but merely as a fine thin
line, apparently a part of the engraving on its surface. Holding the
lower part of the box firmly in his left hand, he turned the rim of the
top slowly about. At once the purpose of this became apparent. Not only
had each pearl, representing a letter of the alphabet, six corresponding
values from rim to center, in any one position, but by turning the rim
around, twenty-six such positions could be secured, making a total of
one hundred and fifty-six different alphabets from which a person
desiring to use a cipher might choose.

Again, however, Duvall was conscious of a feeling of disappointment. One
hundred and fifty-six different ciphers were no better than a single
one, if only one were used. Evidently he had not yet reached the
solution of the