t offset the stretch of the material, but to be on the safe side
and to allow for the notches, add another half-inch for small kites and
more proportionately for large ones. It is advisable to test one pair of
stretchers before cutting another, to reduce the effect of miscalculations.

The stretcher notches should be deep enough to grip the rods well and
prevent them twisting, and one must take care to have those on the same
stretcher exactly in line, otherwise one or other cannot possibly "bed"
properly. A square file is useful for shaping the notches.

Ordinarily stretchers do not tend to fall out, as the wind pressure puts
extra strain on them and keeps them up tight. But to prevent definitely any
movement one may insert screw eyes into the rods near the points at which
the stretchers press on them, and other eyes near the ends of the
stretchers to take string fastenings. These attachments will be found
useful for getting the first pair of stretchers into position, and for
preventing the stretchers getting lost when the kite is rolled up.

The bridle is attached to four eyes screwed into the rods near the tops of
the boxes. (See Fig. 118.) The top and bottom elements of the bridle must
be paired off to the correct length; the top being considerably shorter
than the bottom. All four parts may be attached to a brass ring, and all
should be taut when the ring is pulled on. The exact adjustment must be
found by experiment. In a very high wind it is advisable to shorten the top
of the bridle if you have any doubt as to the strength of your string, to
flatten the angle made by the kite with the wind.

[Illustration: FIG. 115.--Details of stretcher attachment for
diamond-shaped box kites.]

Diamond Box Kites.--In another type of box kite the boxes have four equal
sides, but the boxes are rhombus-shaped, as in Fig. 116, the long diagonal
being square to the wind, and the bridle attached at the front corner.

For particulars of design and construction I am much indebted to Mr. W. H.
Dines, F.R.S., who has used the diamond box kite for his meteorological
experiments to carry registering meteorographs several thousands of feet
into the air.

The longitudinal sticks used at the corners have the section shown in Fig.
115. They are about four times as wide at the front edge, which presses
against the fabric, as at the back, and their depth is about twice the
greater width. This shape makes it easy to attach the shorter stret