508 
Journal of Agricultural Research 
Vol. XI, No. io 
the slab width was thus rendered increasingly eccentric. The position 
of greatest eccentricity was reached at the width of 18.5 feet, in which 
case the load was applied at a point 16 feet from one edge and only 2.5 
feet from the other. Corresponding sections were then split off the 
other side of the slab, reducing the width to 15.5, 11.5, 8.5, and 5 feet, 
respectively, the point of application of the load remaining 2.5 feet from 
one edge. During this stage of the experiment the degree of eccentric¬ 
ity of the load was reduced with each section split off until finally in 
the 5-foot width the load was again applied in the center. In figure 
1 the planes along which the slab was split are indicated by heavy 
dashed lines. 
Complete sets of deformation and deflection readings were made on 
each width of slab, the repeated use of the slab cut to various widths 
being made possible by the fact that the load applied at no time stressed 
the specimen beyond its working stresses. 
As will be noted by reference to figure 1, the concrete and steel deforma¬ 
tion points and the deflection plates were spaced along the middle section 
of the slab parallel to the supports. As this is the dangerous section of a 
slab under such loading, the measurements taken represent the maximum 
deformation and deflection for the loads used at the various transverse 
distances from the point of application. The measurements have been 
plotted to scale, as the ordinates of curves, whose abscissae are the dis¬ 
tances between the points of measurements, and the results are shown 
in figures 2 to io, inclusive. 
RESULTS OF THE TESTS 
As in the tests of centrally loaded slabs previously made, the curves 
of deformation of concrete and steel have been used to determine values of 
the “effective width ” of the slab for the various total widths and positions 
of load. This is done by measuring the areas of the curved with a polar 
planimeter and dividing these areas by the maximum ordinates, the results 
in each case being the value of the effective width corresponding to the 
particular width of slab and position of load. 
These values are shown on the respective curves. This method is 
based on the assumptions that the straight-line theory of fiber-stress 
distribution is applicable to slabs, and that the observed deformations 
are proportional to the extreme fiber stresses in the concrete and steel, 
though it is realized that for various reasons it is impossible to translate 
the deformations into fiber stresses. 
The relation between the computed effective width and the total 
width for the several widths of the single slab tested is shown by the 
solid-line curve in figure 11. As will be noted, it is conservatively 
drawn through the lower edge of the belt of points derived from the 
concrete and steel deformation readings. It shows that the effective 
