36 REPORT OF THE CALIFORNIA EARTHQUAKE COMMISSION. 
a length of about 1,800 feet. There is no evidence that the zone of distributed shear had 
such a breadth in this neighborhood, and moreover 
the displacement of the fence is in the wrong direc- 
tion to be explained by this means. Nor can we refer 
it to elastic rebound as described on pages 17-20 for 
the angle of shear would be more than 1/500 or about 
7 minutes of arc, which is much greater than can be 
allowed. The displacements of the fence are measured 
from its inferred original position supposed to be a 
straight line, but we are not informed how the original 
position was determined. It would not be permissible 
to infer its direction from the continuation of the 
fence outside the eastern stone monument; and if the 
records of the original surveys gives the magnetic direc- 
tion of the line, an imperfect knowledge of the mag- 
netic declination and instrumental errors (if the line 
was run with a compass) would easily account for the 
deviation of 7 minutes between the present line and its 
supposed original direction. It seems probable there- 
fore that the true distortion was confined to a com- 
paratively short length of the fence. There seems no 
clear explanation of the bow-shaped distortion of the 
fence in fig. 34, unless the fence originally had this 
shape. 
The Pilarcitos 30-inch wrought-iron pipe of the 
Spring Valley Water Company runs near the fault for 
a distance of two miles northwest of San Andreas Lake 
and crosses it four times (vol. 1, p. 95). The map 
(fig. 22), taken from the report of Mr. H. Schiissler, 
chief engineer of the company, shows the location of 
this and of some other pipes of the company. Begin- 
iG ning at the northwest the pipe crosses from left to 
Sata eae right at Small Frawley Canyon, and the angle between 
=| 
PLUS 
Pd 

=Soa. 
SH 
ad 4! PIPELINE 

2 
the pipe and the fault-line is 20°; the shortening of 
M the pipe is 7 feet 3 inches, and the offset is 15 inches, 
corresponding to a total shift of 8 feet, as determined 
by the formula on page 34. 
We have no information about the break at the 
next crossing, from right to left, and about a mile 
distant; the pipe runs nearly parallel with, and close 
to, the fault between these crossings and suffered many 
ruptures; in one place it completely collapsed. 
At the next crossing (/), very near the last, the 
pipe crost from left to right at an angle of 15°; it 
was crusht in three places, the total shortening being 
9 feet 8 inches (plate 598 and p. 96, vol. 1); this cor- 
responds to a shift of 10 feet. 
The pipe again crosses the fault near the head of San Andreas Lake, from right to 
left (@), and was pulled apart in two places a total of 6 feet 8 inches (plate 59a); this, 
with an offset of 6 inches, indicates that the angle between the pipe and the fault was 
but 3.5°. 

