downward of 1 in 8, the Bann reservoir--Fig. 8--of Mr. Bateman's design,
where the puddle is 8 ft. broad at the top, and other instances. The same
dimension was adopted for the puddle wall of the Harelaw reservoir, at
Paisley, by Mr. Alexander Leslie, an engineer of considerable experience
in dam construction.

There appears to be a question as to what the composition of puddle
should be, some advocating a considerable admixture of gravel with clay.
There is no doubt that clay intended for puddle should be exposed to the
weather for as long previous to use as possible, and subject to the
action of the air at any rate, of sunshine if there be any, or of frost.
When deposited in the trench, it should be spread in layers of not more
than 6 in. in thickness, cut transversely in both directions, thoroughly
watered, and worked by stamping.

The position of the puddle wall is, as a rule, in the center of the bank
and vertical; but laying a thickness of puddle upon the inner or up
stream slope, say 3 ft. thick, protected by a layer of gravel and
pitching, has been advocated as preventing any portion of the dam from
becoming saturated. There are, however, evident objections to this
method, as the puddle being comparatively unprotected would be more
liable to damage by vermin, such as water rats, etc.; and in case of the
earthwork dam at the back settling, as would certainly be the case,
unless its construction extended over a very lengthened period, the
puddle would be almost certain to become fissured and leaky; in addition,
the comparative amounts of puddle used in this manner, as compared with
the vertical wall, would be so much increased. With the puddle wall in
the position usually adopted, unequal settlement of the bank on either
side is less liable to affect the puddle, being vertical.