It is impossible to produce actual evidence based on losses in the past for every item in the schedule, and many of them-such for instance as the measure of susceptibility to fire and water damage of various goods—were arrived at by canvassing the opinions of underwriters, merchants and manufacturers dealing with the goods themselves. It has been found that wherever the schedule has been applied in cities— and for the matter of that any intelligent schedule or tariff which aims at penalising faults of construction, occupancy, management, etc.-the fire losses

have been reduced.

CHAPTER VIII.

THE CAUSES OF FIRES.

BEFORE passing on to a consideration of the financial principles of fire insurance, I should like to deal to some extent with the causes of fires both in this country and in America, and endeavour to show how profoundly the science of fire insurance is being affected by the changes in modern industrial methods. Lighting and heating methods will at once occur to any one as being most intimately connected with the fire hazard, and the introduction of electricity, both for lighting and heating as well as for power, has worked already something of a revolution. Then again the numerous chemical processes which have grown up have had an important bearing on fire insurance, and managers have been compelled to weigh most carefully their effect on fire risks.

Fire insurance affects every business and the changes and developments of every business react on fire insurance methods. The ideal fire manager would not only be an expert in finance and law but would know the details of every business more intimately than the persons who themselves conduct them. In practice it is impossible for human beings to reach the height of universal knowledge really

necessary, and the proper rating of risks can only be arrived at by pooling the experience in the possession of fire underwriters and their companies. By means of surveyors, who have a special knowledge of the conditions of most businesses, and by the combination of experience, an equitable tariff can be arrived at. But the sole method by which any equitable system of rates can be reached is through a combination or tariff. Private underwriters and non-tariff companies through lack of sufficient data have no means of arriving at the proper rates for each risk and often avail themselves of the labours of the tariff companies and under-cut when they think this can safely be done.

As I have tried to point out more than once already the prevention of fires, though not the primary business of fire insurance companies, is yet a most important secondary issue and one to which their energies are properly directed. There is a small obvious fact about fires which does not receive sufficient attention. There has never yet been a fire, however great, which could not have been put out at an early stage by a bucket of water. The difficulty is so to provide that the bucket of water will be applied at the right time. The fire which destroyed Chicago is stated to have begun in a cattle shed and in a similar way the origin of most great fires can be traced to some small, apparently insignificant outbreak which taken in time would have prevented a disastrous conflagration. No cause is too insignificant to start a fire and once started a whole town may be destroyed. In this country

we rarely experience the devastating conflagrations which are common enough in the United States and in Canada, but even here we have had such outbreaks as the Tooley Street fire and recently the Cripplegate fire, in which the loss has run into a million or more of pounds sterling.

A high insurance authority declares that more than 60 per cent. of the fires which occur arise from preventable causes, and no man can say that his premises are free from the risk of fire however much he may have done towards the exclusion of dangerous lights and heating apparatus. There are scores of ways in which fires may occur apart from human agency, and the public are as a whole profoundly ignorant of the number of ways in which fires may break out. Mr. Moore, a very high authority, states in his work, "Fire Insurance and How to Build," that he has kept for many years a careful record of fires the causes of which were ascertained by adjusters after a careful consideration of all the facts while on the ground. On the business other than farm insurances the fires attributed to unknown causes were about 20 per cent. of the whole number. Exposures to fire risks outside caused about 20 per cent., lightning to buildings 2 per cent., incendiarism internal 4 per cent. and incendiarism external 9 per cent. I believe that the risk from incendiaries is much greater in America, to which this experience relates, than in Great Britain. The remaining causes, 44 per cent., were either directly or indirectly attributed to carelessness. They were matches, smoking, swinging

gas jets and other lighting defects, leaking gas pipes, lamps and stoves, fireplaces, furnaces and other heating apparatus, dryrooms, sparks from locomotives, chimneys, overheating of ashes, spontaneous combustion, explosions, sunlight through glass, rats and mice, friction, gas engines and conflagration. Conflagrations caused from 6 to 10 per cent. of all the losses. While some of these causes are within the control of the occupiers of property, others are outside it altogether. These results were obtained from adjusters who are in a better position to ascertain the true cause of fires than the officials of fire brigades. Out of the total number of fires in New York City in one year nearly a third were due to carelessness with matches and a comparatively small number of these were traced to children.

HEATING.

Now let us come to detail and consider first the more obvious and general causes of fires.

Stoves, stove pipes, etc., are a fertile cause of fires unless the greatest care is taken in regard to their construction and fitting. The pipes should connect with a flue and should never pass out of a window,. through the roof, or through the side of a building.. Pipes which pass through a roof may not only ignite the roof at the point of exit but there is a danger of flying sparks. Chimneys and flues are also dangerous if the brick lining is not sufficiently thick or any woodwork is in direct connection with the brickwork. In the British Standard Fire-Resist