--V--something else; but there is no letter
corresponding to the number 27, so that hypothesis fails. Again, B B B
G is no whole word, nor even the beginning of one; evidently, therefore,
we are not right in that surmise.

"Now let us add together the first and second pair of figures in every
group; for it is only by testing every possible combination of these
exasperating figures that we shall arrive at their meaning. By doing
this we get 4 and 9, which correspond to D and I. Now that looks more
promising, so let us take the next group 1819. These, added, make 9 and
10, corresponding to I and J. This gives us D I I J. That again,
Harry, does not seem to mean very much, does it?"

"No," replied Harry, "it certainly does not. Still, let us go on; we
may make something out of it yet. The next group is 1919, which makes
10 and 10 or J J; and the next group makes 8 and 4, or H and D.

"Now let us put all these together. By doing so, we get D I I J J H D,
which certainly does not look like any language. We can make no words
out of those letters, whichever way we arrange them, so it seems that we
are wrong again in our method."

"Never mind, my friend," said Roger, "let us still go on; it will not do
to be discouraged so soon. There certainly is some translation to that
mass of figures, I feel certain, and I am determined to find it. Now,
how can we go about it next? I have it! Let us take each group as
representing one letter instead of two or four, as we did before. What
shall we then get?

"We now have 13, 19, 20, 12, 11, 12, 13, 19 for our first line,
representing, in letters, M S T L K L M S.

"This, again, conveys no meaning; nor can any words be formed whichever
way we arrange the letters.

"Now, instead of adding each figure separately, let us add each set of
two, that is, 22 and 27 and 18 and 19, then 19 and 19, and so on, and
see what we get then."

"But," objected Harry, "if we do that, we shall get numbers which have
no corresponding letters. I mean that by doing as you suggest we should
have 49, 37, 38, and our numbers corresponding to letters only go up as
far as 26, which stands for Z, so that method cannot be right. It seems
to me that this thing is beyond us, Roger; I do not see what we can make
of such an awful jumble of figures."

"Gently, my lad, gently," remonstrated Roger, "we will not give it up;
we may as well be worrying over this cryptogram as doing nothing, and
better, be