14 R. T. CHAMBERLIN AND W. Z. MILLER 
Troublesome difficulties arose from the use of these materials. 
Owing to the incoherence of the sand, the competent plaster 
layers, instead of rising along a single fault plane after fracturing 
had occurred, were thrust into the sand layers, producing a dove- 
tail effect. Obviously it was necessary to add something to the 
sand to correct this and to give the sand more coherence, and yet 
at the same time the sand layers were to be kept relatively incom- 
petent. With this end in view varying amounts of plaster were 
added to the sand and the result noted. But it was found that 
if sufficient plaster were added to furnish the desirable coherency, 
the layer became too competent for the purpose of the experiment. 
Clay alone was tried, but owing to the weakness of the clay, or its 
lack of adhesion to the plaster, dovetailing again resulted. To 
obviate this, plaster was added to the clay in the proportion of 
about two parts of clay, with some sand, to one part of plaster. 
This combination was successful, though in different experiments, 
where varying degrees of competency were desired, somewhat 
different proportions were used. As would be expected, the plaster 
layers carried the thrust, and the whole mass acted as a rigid 
body until these competent layers were fractured. The influence 
of softer layers after fracturing will be discussed later. 
In certain experiments it was desired to give the competent 
layers somewhat greater plasticity. This necessitated increasing 
the plasticity of the incompetent layers as well, in order to keep 
the relative competency the same. For this series of experiments 
a mixture of plaster, clay, and sand was used for the competent 
beds, varying in proportion from three parts of plaster and one of 
clay and sand to two parts of plaster and one of combined clay 
and sand. For the incompetent beds either pure clay was used, 
or a mixture of clay and plaster varying in proportion from four of 
clay and one of plaster to two of clay and one of plaster. Such 
strata required from three to seven days to harden, depending upon 
the proportions of clay and plaster used. 
RELATION OF FAULTS TO STRESS AND STRAIN 
Stresses may be defined as the forces developed within differ- 
ent parts of a structure under the action of external forces operat- 
