square to square,
without lifting them or removing them from the box, until they are
placed in their natural order. It is easy enough to move the squares
up to 12; but to get the last three into order is often a puzzle
indeed. If the figures fall in either of the following positions--13,
15, 14; 14, 13, 15; or 15, 14, 13--the problem is unsolvable; it
follows, therefore, that the last row must be either 14, 15, 13; or
15, 13, 14. If you get the cubes into either of these positions, you
can easily bring them right; but if you cannot, the only way is to
begin the game all over again. Several other ways are suggested.
Cavendish (Mr. H. Jones) thinks he solves the puzzle by turning the
box half round; but as this is only possible when the figures are on
circular pieces of wood, his solution merely cuts the knot, instead of
unravelling it.

2592. The Thirty-Four Puzzle.

This is an adaptation of tho old magic square, which amused the
philosophers of old. A sketch of it appears in Albert Durer's painting
of Melancholia. Sixteen discs or squares, numbered from 1 to 16, are
placed indifferently on the table--or they may be in the fifteen box;
and the puzzle is to so arrange them as to make the sum of the figures
add up to 34, whether counted up, down, across or angularly. Here is
the solution:

--------------------------- ---------------------------
| | | | | | | | | |
| 1 | 15 | 14 | 4 | | 1 | 8 | 13 | 12 |
| | | | | | | | | |
|------+------+------+------| |------+------+------+------|
| | | | | | | | | |
| 12 | 6 | 7 | 9 | | 14 | 11 | 2 | 7 |
| | | | | | | | | |
|------+------+------+------| |------+------+------+------|
| | | | | | | | | |
| 8 | 10 | 11 | 5 | | 4 | 5 | 16 | 9 |
| | | | | | | | | |
|------+------+------+------| |------+------+------+------|
| | | | | | | | | |
| 13 | 3 | 2 | 16 | | 15 | 10 | 3 | 6 |
| | | | | |