a uniform loss of metal over the surface, but it comes out in small particles, leaving an irregular surface. The better physically the quality of the steel the smaller are the particles, and the smoother the surface, and the longer the steel wears.

Under the wheels of the cars and drivers of the locomotives, the pressure per square inch of the surfaces in contact in most cases exceeds the elastic limit of the steel, measured by the usual method of tension. The limit rises on the surface of the rails and tires, but is still insufficient to prevent rapid flow and loss of the small irregular surfaces in contact.

The slow rate of wear of the light-headed rails under the then existing wheel tonnage was studied, and deeper heads given to subsequent rails, for supposed increased service. Experience now shows that rails do not wear down in a smooth, uniform manner, but, on the contrary, the wear is very uneven per length of rail, and they are removed from the track on that account before they are fully worn

out.

What is needed today for present, and to anticipate the average, increase of wheel tonnage is to widen the head of the rail, thus increasing the surface of contact, in order to distribute the weight over larger areas, and thereby check the rapid loss of metal.

To check the deflections, stiffer sections of rails are needed, or in other words, the material must be distributed so as not only to decrease the rate of wear, but make a stronger rail. The problem is not wholly one of weight of rails, but also of form.

The heaviest rail in use now is eighty pounds to the yard. The section as used by the New York Central & Hudson River Railroad is five inches high and nearly the same width of base, the head is 2 inches wide but comparatively shallow. The upper corners of the head are 5-16 of an inch radius, while that of the top is 12 inches, giving a broad bearing surface for the wheels, thereby checking the rapid increase of wear. The increased wear of the treads of all the wheels is of far more importance than the small percentage of wheels condemned for sharp flanges.

The 72-pound rail which has been in use on the Boston & Albany Railroad since 1880 has been brought to such good surface in the track that only from five to eight deflections in length of 11 feet under the weight of a 34-ton car exceeded one-eighth of an inch per mile.

The speaker closed with a description of his "Dynagraph and Track-Inspection Car." By means of this car an accurate record of the condition of the track at the time the car passes over it is made on paper; besides this, it throws paint under the head of the rail at every point where the deflection in 11 feet is more than a certain amount, which can be adjusted, thus telling the section men the exact places which need attention. The first time a car is run over a road it is usual to paint all points where the deflection is more than five-sixteenths of an inch, but on subsequent trips this is reduced sometimes as low as one-eighth of an inch, and in one case to three thirtyseconds of an inch. This is about as close as it is possible for the section men to keep the track.

Diagrams as they came from the car, taken from several roads, were exhibited.

A large number of lantern views were projected on the screen to illustrate various features of the subject.

The meeting closed with a vote of thanks to the speaker for his very interesting and instructive paper.

MEETING 362.

The Martin-Wilson Automatic Fire Alarm.

BY MESSRS. A. H. KENDALL AND M. MARTIN.

The 362nd and annual meeting of the SOCIETY OF ARTS was held at the Institute on Thursday, May 12th, at 8 P. M., Mr. C. J. H. Woodbury in the chair.

After the reading of the minutes of the previous meeting, and the election of new members, the Nominating Committee presented their report, and officers were elected for the ensuing year.

The report of the Executive Committee was read, and ordered placed upon the records.

The Permanent Meteorological Committee then reported through its chairman, Prof. W. H. Niles.

REPORT OF THE METEOROLOGICAL COMMITTEE.

Prof. NILES said that the duties of the committee during the year had been unusually few. The Signal Station in Boston had been inspected by the members, both collectively and individually. It was found to be well kept, the instruments appeared to be in good condition, but the sergeants had not at all times been aided by competent assistants.

By the failure of Congress to make the requisite appropriations, the Signal Service had been forced to suspend some of the most important functions of the Weather Bureau. The Service had not been able to transmit by telegraph the reports of the weather at various important points, hence the issuing, at Boston, of a daily weather map had been rendered impossible for a time. Such interruptions seriously reduce the value of the Signal Service to the country. It is much to be desired that the Weather Bureau should receive that financial support which shall enable it to prosecute its work effectively and continuously.

The published "Indications" of the weather had received during the year more adverse criticism than usual. In many instances the criticisms did not accord with the facts, but there have been good reasons for the belief that, in a broad country like our own, weather warnings can be made more accurate, and therefore more valuable, than ours have been during the past year. If there is not an improvement in the "Indications," it may be well for the members of the Society to express their united request that a weather service should be so sustained by stated and adequate appropriations that the best efforts of the most competent men may be continuously employed in work which shall yield the greatest benefit to commerce, agriculture, and other industries.

THE MARTIN-WILSON AUTOMATIC FIRE ALARM.

The chairman then introduced Mr. A. H. Kendall, who read a paper descriptive of the "Principles of the Martin-Wilson Automatic Fire Alarm."

Mr. KENDALL said The sharp competition of today is forcing business into conditions where it can obtain only reduced profits, which formerly would not have been considered to be an equitable remuneration for the capital invested. And these sparse profits in commercial affairs and manufacturing enterprises are now made out of what was wasted a few years ago. One of these items is the fire waste as represented in the cost of insurance, which is of course strictly based on the actual destruction of fires. The general consideration of insurance a few years ago was treated as an act of fate. There was but little elasticity in the matter of insurance rates, and if any economy was exercised, it was in the reduction of the amount of insurance rather than seeking to diminish the rate, and still preserve an amount of insurance adequate to indemnify the assured to an equitable degree in case of fire. The systematic effort to reduce the rate of insurance by improved construction and fire protection is one of the most modern features of business. The aggregation of values in our business centers, with large buildings closely crowded together, and the augmented fire hazards from these conditions, as well as from the greater speed of machinery, and the introduction of many processes in industrial chemistry, such as drying at high temperatures, and the wide use of artificial light, have increased the risk of fire to an extent far beyond the conditions of less than a generation ago.

At the same time, the defences against fire have been increased. Our fire departments have been developed until, both in organization and equipment, they resemble a standing army. The power of the law is exercised in the same direction, limiting the quality of illuminating oil, its methods of storage and sale, and in cities the construction of buildings to the utmost detail practicable with enforcement. But, nevertheless, the terrible ravages of fire show that these precautions have not kept pace with the increasing conditions of hazard, and I appear before you this evening to offer a few words in explanation of the latest application of modern science to this problem as represented by the Martin-Wilson automatic fire alarm, its function being that of an ever-vigilant custodian by its instantaneous operation to sound an alarm which will save the precious seconds at the beginning of every fire, and summon the assistance of the fire department and other help at the earliest moment. Captain John S. Damrell, for many years the chief of the Boston fire department, and now at

the head of the bureau of building inspection, once uttered a great truth in saying that a great fire was a neglected small one.

I purpose submitting some facts relative to thermostats in general, and their application to the Martin-Wilson automatic fire alarm system in particular, with a comparison of the different systems of application. It is useless at this late day, when the losses by fire in the United States are so heavy that fire insurance has almost ceased to be a profitable business, to go into the question of the value of an automatic fire alarm. Any invention or application that is absolutely certain to automatically indicate the location of a fire in a building in its incipiency before it has had time to do much damage, and when it can be easily extinguished, deserves the earnest attention of insurance companies, property owners, capitalists, and the general public.

The foundation of an automatic fire alarm is the heat detector, and this company claim that their heat detectors or thermostats are at once the most scientific, the simplest, and the most reliable instruments yet invented; and that their system of electrical circuits and apparatus used in connection therewith, for the purpose of sounding a fire alarm by the heat of the fire itself, cannot be excelled. The idea of utilizing electricity to give notice of undue heat is by no means a new one. But the great difficulty has hitherto been in its execution, from the fact that a fire being of comparatively infrequent occurrence, the circuits, by reason of an accidental break, earth connection, or a weak battery, would be apt to be in an inoperative condition when a fire occurred, and so fail to do the work expected of them. In 1830, Dr. Ure, the celebrated English chemist, invented a thermometer in which the movement of a spring composed of two unequally expansible substances, for instance brass and steel, indicated the temperature to which it was exposed. This instrument he called a thermostat, and he utilized its motion to regulate the valves or dampers of furnaces, and since that time thermostats and thermometers embodying the same principle have been devised by different inventors, a familiar example being the common metallic thermometer. From 1830 to 1881 numerous thermostatic heat detectors were invented, based on various principles, such as the expansion of metals and air, beeswax, and similar substances, and the rending or rupturing of fragile vessels by the vaporization of volatile liquids, such as alcohol, ether, etc.