304 REPORT OF THE CALIFORNIA EARTHQUAKE COMMISSION. 
With regard to the scale on which the experiment was performed, the question naturally 
occurs: Would similar results have been obtained if the car were very much largei ? 
One can not be certain, but it seems that such would have been the case if the frequency 
and amplitude of the car’s motion were the same. Several experiments were performed 
with a depth of 15 em. and 25 cm., and the results were substantially the same. The 
car was also divided by partitions running at right angles to the direction of motion, 
making a compartment of only half the length. The results tabulated below show that, 
at least in this case, the motion of the block embedded in the sand was not greatly affected 
by the presence of the partitions. 

Length of car between partitions in em. |101 49 101 101 49 49 101 
Frequency in double vibrations per sec. 2.2 2.25 2.3 2a 2.4 Dele 2.25 
Amplitude of carinem. . ... . 7.5 7.6 TH 7.6 7.5 7.5 7.5 
Amplitude of block in sandinem. . . NSS, 9.3 11.3 10.8 10.8 11.4 11.8 

In the experiment just described the material used contained 20 per cent water; with 
less water the partitions would doubtless have a greater effect in restricting the motion. 
It is also probable that with a larger car the relative motion of car and load would have 
been greater. 
Another question which is likely to oecur is: Does the solid or semi-fluid mass 
with which the car is loaded have a free period of its own which is comparable with the 
vibrations imprest upon it by the to-and-fro motion of the car? To elucidate this matter, 
the car was partially filled with water and the free period of gravitational waves deter- 
mined experimentally. The frequency of such waves was found to be 1.06. However, 
the load, instead of being like water, was in all cases exceedingly viscous and plastic. 
This condition would in any case decrease the natural vibration frequency of the load, 
and in the present case the viscosity was so great that the load could not possibly have 
any vibration independently of the oscillatory motion of the ear. 
Finally the question may be asked: What is the explanation of the fact that the load 
on the car (or the major part of it) oscillates thru a greater amplitude than the car which 
causes the motion? At present I have no comprehensive explanation of this fact. It 
undoubtedly depends upon the inertia of the load, combined with the greater or less 
freedom with which it yields to imprest forces. The load in the car is set into motion 
by two sets of forees: (1) On account of the motion of the bottom of the car a tangen- 
tial force is exerted on the bottom of the load and this is transmitted upwards by the 
rigidity of the load, or, exprest otherwise, by the mutual friction of successive layers of 
the load. (2) On aécount of the advancing motion of the end of the car the load re- 
ceives a thrust which is transmitted thru the material by its resistance to compression. 
Sometimes one of these sets of forces is of greater importance, and sometimes the other. 
One would be apt to think the end thrust was of the greater importance, but this is 
certainly not always the case, for when the load consists of a mound not resting against 
cither end of the car, the block embedded in the top of the mound may oscillate with 
a much greater amplitude than the car. (This was experimentally demonstrated.) In 
this case there can be no end thrust whatever. In some cases the end thrust may be 
more effective than the tangential force; this is probably the case when the frequency of 
motion is rather high. ) 
To those interested in seismology the important question is: How do these experi- 
ments help to explain the greater destructiveness of earthquakes in regions where foun- 
dations are in alluvial soil than where foundations rest directly upon rocky strata? To 
pass from experiments upon a box containing half a ton of soil to the destructive effects 
of an earthquake is certainly a great leap. In taking such a step, one is very likely to 
make mistakes. However, it seems to me beyond question that a soft, semi-fluid mass of 
