he surrounding concrete? The greatest stress will come
on the rod at the point where it leaves the concrete, where it is a
maximum, and it will decrease from that point inward until the total
stress in the steel has been distributed to the surrounding concrete. At
that point the rod will only be stressed back for a distance equal in
length to 50 diameters, no matter how far beyond that length the rod may
extend.

The distribution of the stress from the steel rod to the concrete can be
represented by a cone, the base of which is at the outer face of the
block, as the stresses will be zero at a point 50 diameters back, and
will increase in a certain ratio out toward the face of the block, and
will also, at all intermediate points, decrease radially outward from
the rod.

The intensity of the maximum stress exerted on the concrete is
represented by the shaded area in Fig. 10, the ordinates, measured
perpendicularly to the rod, indicating the maximum resistance offered by
the concrete at any point.

If the concrete had a constant modulus of elasticity under varying
stress, and if the two materials had the same modulus, the stress
triangle would be bounded by straight lines (shown as dotted lines in
Fig. 10); but as this is not true, the variable moduli will modify the
stress triangle in a manner which will tend to make the boundary lines
resemble parabolic curves.

A triangle thus constructed will represent by scale the intensity of the
stress in the concrete, and if the ordinates indicate stresses greater
than that which the concrete will stand, a portion will be destroyed,
broken off, and nothing more serious will happen than that this stress
triangle will adjust itself, and grip the rod farther back. This process
keeps on until the end of the rod has been reached, when the triangle
will assume a much greater maximum depth as it shortens; or, in other
words, the disintegration of the concrete will take place here very
rapidly, and the rod will be pulled out.

In the author's fourth point he belittles the use of shear rods, and
states: "No hint is given as to whether these bars are in shear or in
tension." As a matter of fact, they are neither in shear nor wholly in
tension, they are simply in bending between the centers of the
compressive resultants, as indicated in Fig. 12, and are, besides,
stressed slightly in tension between these two points.

[Illustration: FIG. 11.]

In Fig. 10 the stress triangle indicates