DIRECTIONS OF VIBRATORY MOVEMENT. 307 
did not always insure their safety. Many that stood were found not to be built up from 
the ground, but to rest on shelves somewhere within the house. This method of build- 
ing seemed to preserve the chimney intact in the majority of cases. A few chimneys 
owe their preservation to their low, solid structure above the roof ; many did not fall 
because they were well-braced, either by being inclosed in a wooden casing or a coating 
of cement, or by being held by iron rods clamped into the brick. A striking example 
of the advantage of an iron rod as support was that of a 2-story house in San Mateo. 
This house had a brick foundation and a slender chimney 14 fect high, supported by an 
iron rod. The chimney stood perfectly. 
A great many chimneys that stood well above the roof were badly damaged at the 
base or within the house, and many were cracked above the roof and shifted a short 
distance horizontally. The use of cement in the mortar saved the chimneys in some 
instances, but a common effect of the shock on chimneys so built was to crack them 
somewhere and make them fall in one piece. In this way solid masses of great weight 
were sometimes pitched on to roofs and other parts of buildings, and the result was 
much greater damage to the house than was caused by chimneys built with lime mortar. 
Chimneys laid with lime mortar generally broke in many pieces or fell as loose bricks. 
The use of cement below the roof was apparently helpful, as the chief danger to that 
part of the chimney is from cracking rather than from falling, and the cement is much 
less apt to crack than the lime. The use of lime mortar above the roof is better, unless 
the chimney is to be boxed and braced. The construction of boxes around chimney 
tops, and the bracing with iron rods, are two simple and efficient preventives to the 
falling of chimneys of which comparatively few have made use. 
Chimneys other than brick. — Many of the small houses of San Mateo County use terra- 
cotta thimbles or chimney pots, in place of brick chimneys. Their efficiency against 
earthquakes is conclusively shown by the fact that a large proportion of them stood 
unhurt, even when built in several sections. From 90 to 95 per cent of these chimneys 
past through the earthquake without harm. Galvanized-iron pipes, and stove pipes 
used as chimneys, were likewise unhurt in most cases. The few chimneys that were 
built entirely of concrete proved to be much stronger than those of brick. 
Plaster. — In almost all houses with plastered walls, the plaster was cracked more or 
less seriously or broken off in sheets. The plaster or stucco on the outside of houses 
was badly damaged. In the majority of the houses, some of the walls — usually not all — 
were seamed with small cracks which ran in every direction and frequently in lines 
parallel with the laths. In other cases the cracks were wide and the walls were in large 
part laid bare. 
The second table on page 365 gives the statistics regarding the cracking of the plaster. 
The first column includes the cases in which the plaster was almost unhurt or only 
slightly cracked. Most of these buildings did not require replastering. The second 
and third groups include the buildings more seriously damaged. Replastering was 
necessary in the second and third groups. The plaster on the ceilings of houses was 
much less affected than that on the side walls, and in the majority of cases was unhurt. 
In 2-story houses the plaster was rarely damaged as severely on the second floor as 
on the first floor, and in wooden houses of three stories it was often observed that the 
plaster on the third floor was uninjured. This restriction of the damage to the ground 
floor may be due to the breaking of the plaster by short, sharp movements near the 
ground, which were translated above into the swaying of the entire upper story. That 
the plaster did not crack much on ceilings was probably due to the fact that the ceilings 
(and the floors above) were not subjected to so much strain because they moved as one 
piece. Thick coatings and varieties of hard plaster seem to have been less damaged. 
New plaster not yet dry was not affected in the few cases observed. 
