ON THE EARTHQUAKE PHENOMENA OF JAPAN. 373 
ing attached to the moving ground. The extent to which the acquisition 
of momentum may be avoided by the adoption of one or all of these 
methods may be judged of by reference to the three previous sections of 
this report. To resist the effects of momentum which cannot be cut off a 
building, and therefore tends to shatter it, the following rules appear to 
be among the most important :— 
1. Bear in mind the fact that it is chiefly stresses and strains which 
are applied horizontally to a building which have to be encountered. 
Ordinary masonry arches are continually cracked by earthquakes, inas- 
much as they are only designed to resist the stresses due to gravity. 
Tron or wooden lintels, or at least tie-rods, may be used to strengthen or 
replace such arches. Arches which meet these abutments at an angle 
are more liable to be cracked than those which meet them with a curve. 
A vertical line of openings like doors or windows in a building constitute 
a vertical line of weakness to horizontally-applied forces. 
2. Avoid coupling together two portions of a building which have 
different vibrational periods, or which from their position are not likely 
to synchronise in their motion. If such parts of a building must of 
necessity be joined, let them be so joined that the connecting link will 
force them to vibrate as a whole, and yet resist fracture. 
Brick chimneys in contact with the framing of a wooden roof are apt 
to be shorn off at the point where they pass through the roof. Light 
archways connecting heavy piers will be cracked at the crown. To ob- 
yiate destruction due to these causes a system of construction similar to 
that to be seen in several of the buildings of San Francisco, Tokio, and 
Yokohama may be adopted. This essentially consists of tieing the build- 
ing together at each floor with iron and steel tie-rods crossing each other 
from back to front and from side to side. In connection with this subject 
I may mention that experiments have shown that in strong earthquakes 
the abutments of heavy archways approach and recede from each 
other, cracks in buildings open and shut, other cracks grow longer, and 
finally that two points of ground not ten feet apart do not synchronise 
in their motion. These facts show that the assumption of a difference 
in vibrational phase in parts of a building shaken by an earthquake is 
not simply an assumption made to explain certain often-repeated obser- 
vations. 
3. Keep the centre of inertia of a building or its parts as low as pos- 
sible. Heavy tops to chimneys, heavy copings, and balustrades on walls 
and towers, heavy roofs and the like are all sources of danger to the por- 
tion of the structure by which they are supported. When the lower part 
of a building is moved, the upper part by its inertia tending to remain 
behind often results in serious fractures. All the chimneys in Tokio 
‘and Yokohama which have fallen in consequence of their ornamental 
heads have been replaced by shorter and thicker chimneys without the 
usual coping. The roof of a portion of the Engineering College rests 
loosely on its walls, and has therefore a certain freedom. In Manila many 
heavy roofs have been replaced by roofs of sheet iron. Walls may be 
lightened in their upper parts by the use of hollow bricks. Such ver- 
tical motion as may exist is also partly obviated by light superstructures. 
Vertically-placed iron tie-rods give additional security. 
If these and other rules which are the result of experiment and ob- 
servation could be adopted in earthquake countries, it is certain that the 
loss of life and property might be greatly diminished. Before building 
