The present earthquake swarm is no indication. Such earthquake swarms are seldom followed by great earthquakes. When they the epicentre of the main earthquake rarely coincides with that of the swarm (Davison, 1936, 246-264). It is, therefore, obviously useless to move away a few miles thinking that safety will be so obtained. For this reason I did my best to discourage those who wished to run away from Paliyad. I advised them to live in tents away from walls of houses till the earthquake shocks should abate and then to return to Paliyad and rebuild or repair the damaged buildings.

PART II.-EARTHQUAKE DAMAGE AT PALIYAD.

The damage done by the earthquakes in Kathiawar up to the time I left on the 1st of August was confined to the town of Paliyad. Here most of the pukka buildings have been cracked, but none have fallen. The kutcha buildings are also cracked, especially at the corners, and a few have fallen.

I am not competent to assess the damage done, but probably it lies somewhere between the figures of Rs. 30,000 and Rs. 3,00,000; the former figure if cracked buildings can be repaired without demolition, the latter if they have to be pulled down. Now each damaged building presents its own engineering problem, and really requires to be examined by an engineer with experience of repair of earthquake damage in other parts of India.

The Chief Engineer, Roads and Buildings, P. W. D., Bengal, has very kindly supplied me with the following notes from P. W. D. practice in Bengal.

"Earthquake cracks are treated much like cracks from settlement or any other cause. If walls are not out of the plumb the cracks are opened out well, and then grouted Note on repairing with a mixture of 2 parts of sand to 1 of earthquake shocks. cement. Where the crack is at all wide it is good practice to push small pieces of fresh rock into the cement concrete filling the crack. In bad cracks tell-tales' of cement are usually inserted so that any further movement of the crack can be observed. Steel tie-rods can also be used in cases where this is considered necessary. The usual practice is to clear a hole on either. side of the crack for the tie-beam. The space around the steel tie-rod is then filled in with concrete. Walls well grouted or tied together

in this manner are considered to be as strong as they were before the earthquake damaged them.

Masonry walls which are out of the plumb should generally be dismantled."

New buildings.

It is the practice in parts of the earth specially liable to earthquakes to build the larger structures with braced steel frameworks, or of reinforced concrete, and the smaller buildings of light and elastic materials such as bamboo and thatch. Details of such earthquake-proof structures can be obtained in a chapter on the design of earthquake-resisting buildings (Freeman, pp. 795-820). References to the best literature on this subject are given in the same book (pp. 709-724). Construction of this kind would be expensive in parts of Kathiawar at a distance from the ports, as none of the necessary materials would be readily available, nor could the expense be generally justified. I have shown (p. 584) that the danger from earthquakes in Kathiawar is only a small fraction of the danger in Baluchistan, and yet few buildings in that country outside Quetta have been specially designed to withstand earthquakes.

While rejecting as impracticable any building policy aiming at making Kathiawar earthquake-proof, I strongly favour such modifications of the present building methods as will tend to reduce damage from earthquakes.

In the part of Kathiawar visited by me the pucca buildings Existing building pracare made either of random rubble or ashlar, tice in Kathiawar. the former being the commoner.

The stone used for the rubble is usually a heavy black basalt, that for the ashlar may be one or other of the sandstones so common in northern Kathiawar, or it may be a limestone known locally as Porbandar stone. Lime mortar is used in most cases, but cement is seen to some extent along the railway lines. The foundations of pucca buildings in all central Kathiawar are on solid rock.

Heavy masonry walls such as those described are more vulnerable to earthquakes than other forms of building construction, but they are cheap and cool and stand a long while. I do not consider it necessary or desirable to replace them by earthquake-proof structures, but more care should be taken in building new houses to see that the mortar used is the best available, and that the stones are

set with a good bond. The cracks in Paliyad were in most cases worst where least attention had been paid to bond. The few pucca walls which had fallen had all been built with mortar so poor that it could be crumbled between one's finger and thumb. To use such poor stuff in a country like Kathiawar, where excellent lime is readily available, is almost criminal.

Design of Buildings.

As regards the general design of buildings the following remarks in Messrs. Auden and Ghosh's report on the 1934 earthquake in Bihar are equally applicable to Kathiawar.

"Buildings should be of simple design, the parts so well-tied and the whole structure so rigid that it would react as one unit to earthquake waves. The several parts of irregular buildings do not synchronise during a shock, and severe stresses are set up between the individual units. These remarks apply especially to high structures, e.g., a church steeple erected on a tower is frequently destroyed from the base of the steeple; the tower and steeple each having a different vibration period and opposing directions of movement, with the result that the acceleration can be doubled."

"Buildings should be kept as low as is conveniently possible compared with their lateral dimensions. The object should be to keep the amplitude of vibration as low as possible at the highest part of the structure, and so reduce acceleration, for any particular period of vibration."

"Buildings of irregular shapes, with wings, protruding verandahs, porches, etc., have invariably suffered. The same applies to buildings to which additions have been made by the abutting of new walls directly, on earlier ones, without dovetailing. The whole building should form one unit. unit. Verandahs and porches should not consist of a series of independent pillars with a roof resting on top, but should be integral parts of the building."

"Excrescences such as towers, turrets, pinnacles, etc., are dangerous both to human beings and to buildings, and should be avoided. Flimsy parapets, balustrades, and similar structures have caused death to many."

"Windows are a source of weakness to buildings, and a more careful and better spacing should be attempted. Windows should

be kept away from outer corners of buildings as far as possible. Wide window areas should be compensated by stronger intervening walls.”

"The use of timber-frames in kutcha-pucca and kutcha buildings should be encouraged. Buildings consisting of a platform or soleplate supporting timber verticals that are well-tied by cross and diagonal beams and firmly attached to the roof are preferable. The weight of the roof would thus be shared both by the walls and the pillars and the tendency of the walls of such a building would be to move together as a unit. Timber pillars are greatly to be preferred to brick pillars."

There are some very old houses in Paliyad made of masonry with a wooden framework. These were not damaged at all by the recent earthquake. I have no doubt that an extension of this form of building in Kathiawar would be a valuable insurance against earthquakes, but it is doubtful whether suitable timber for this purpose could be obtained now-a-days at a reasonable cost.

The remarks above about buildings being kept low compared to their lateral dimensions are particularly applicable in Paliyad. Many of the shopkeepers' houses are high and narrow. The upper storeys have in most cases been badly cracked. The mortuary attached to the hospital is a particularly good example. There is only a single very small room, but it is about 15 feet high and has a heavy roof. Under the impact of the earthquake it must have vibrated very severely, with the result that it has been badly cracked, though it is a new and comparatively well-built building.

According to the Chief Engineer, Roads and Buildings, Bengal, it is good practice in earthquake areas to put in a ring of cement concrete six inches thick and the full width of the wall just above the level of the lintel all round a house. This costs little, is valuable in stopping cracks, and generally strengthens the building at its weakest point. Where there is more than one storey a ring is needed for each storey.

Roofing.

Lightness is the object to be aimed at in roofing. The Mangalore and country tiles now used for the purpose are rather heavy. In a severe earthquake they might crash down as one piece, burying everything under them. The large sheets of asbestos now sold for roofing would be much lighter and less destructive if they fell. They are said to be little dearer than Mangalore tiles. If they could

gradually replace the tiles throughout Kathiawar it would greatly reduce the danger from earthquakes.

Mud Huts.

The homes of the cultivators have mud walls. Where these have been built on a good wide stone plinth, and with a suitable taper, they have not collapsed. Those walls that have fallen were damp, the moisture having risen and permeated them through a badly constructed and inadequate plinth. If these plinths had watertight lime fillings, it would be a great improvement.

It is the local practice to add straw, pieces of tile, or pottery, and small chips of stone to the mud walls. This practice is probably excellent, but in some of the mud walls which had fallen I saw quite large rounded boulders. These are undoubtedly a source of danger, and should be avoided in this form of building.

Nearly all the mud walls were cracked at the corners. Cracking of this sort would be greatly reduced if there is a timber framework such as has already been described (p. 588). It is doubtful, however, if this would be worth the expense.

Most of the mud huts have roofs made of half-round country tiles. These would certainly crash down and bury the inhabitants in a bad earthquake. I regard them as much more dangerous than the mud walls, which would probably collapse without doing much harm. Thatch or asbestos sheeting would be a great improvement. Unfortunately the former is not available in this part of Kathiawar and the latter is probably too expensive. The use of these materials for roofing should be encouraged as far as possible.

Width of streets.

A feature emphasised in the recent report on the Quetta earthquake (West, p. 224) is the loss of life due to the narrowness of the streets. This 'made it impossible for people who did have time to escape out of their houses to reach a place of safety'. West recommended that the width of the streets should be not less than the combined height of the houses on either side' (p. 235).

The streets in Paliyad and in most of the older towns in Kathiawar are narrow. Every opportunity should be taken to increase their width. The coastal towns in Kathiawar, especially those in the