366 REPORT OF THE CALIFORNIA EARTHQUAKE COMMISSION. 
The testimony is good in all cases that structures on the hills suffered less severely 
from the earthquake than those on the plain. If a large amount of similar data could 
be collected on the low, alluvial, often marshy, flat land bordering the bay, it would 
probably be shown that the movement there was still more intense. Houses, however, 
are not frequent there. In low bottom-land there were indications of great intensity, 
and especially in the case of ground artificially filled in. A good example was given by the 
electric railroad track a few miles north of San Mateo, shown in plate 97c, p. It was 
built over the low land on a heavy, but loose, embankment of earth and stone. At one 
place this roadbed was shaken apart between the rails, and a crack from 1 to 2 feet 
wide and extending down many feet, nearly if not quite to the level of the valley, was 
formed in it for a distance of over 1,000 feet. It ran northwest and southeast, parallel 
with the road, and thruout that stretch not one of the heavy steel rails was left unbent. 
One 30-foot rail that was examined had been bent 2 feet horizontally and 10 inches ver- 
tically. Such wrecking of railroad tracks occurred wherever the underlying founda- 
tion was loose, but the stretches of track on solid ground were not affected. The 
low, muddy land along San Francisco Bay, east of San Mateo, was seamed with cracks 
by the earthquake. 
CONCLUSION. 
The following are the main conclusions arrived at in the course of the work: 
1. It is evident that much of the damage to houses, as well as to their contents, could 
be avoided by judicious construction. The disadvantages of certain classes of structure 
should be acknowledged, and search made for more successful styles. Houses prac- 
tically earthquake-proof can be built easily and cheaply. 
2. The dominant directions taken by moving bodies during the course of the earth- 
quake shock were southwest and northwest, with movements northeast and southeast 
only second in number. There appear to have been felt in this region two main thrusts 
or sets of movements that emanated from the fault-line in southeast and northeast 
directions. 
3. The shock was less heavily felt on the hills than on the level land. The lower 
slopes were affected in an intermediate degree. The difference in the two extremes was 
probably almost as much as one degree of intensity in an earthquake table of 10 units. 
