well to see that this new step is justified, and it can only be justified by a practically perfect continuity of service.

Many railroads electrifying will insist upon the internal control of their entire power, others will mingle purchased power with their own power, and, finally, still others will depend entirely upon purchased power. I see no particular difficulty arising in any one of these three arrangements of supply, and I have agreed with Mr. Torchio that the only true basis for a contract by a railroad for the supply of outside power will be that that power can be supplied with as equal reliability as the power could be produced by the railroad company and at equal cost.

The higher thermal efficiency secured in the use of large generating units affords an opportunity for large central-station plants to sell power to railroads, a measure of the justifiable profit in this sale being in the difference in efficiency between the smaller generating station (required by the railroad) and that of the central station. Such an arrangement as Mr. Torchio has pointed out provides an economical supply of power, and at the same time obviates the necessity of the railroad company investing in this feature required for electrification. Indeed, and digressing for a moment, I will go a step further under Mr. Torchio's thought and say that in these days when railroads are so hard up for cash the "Equipment Trust" offers a relief in the matter of the purchase of electric motive-power equipment, leaving as the only cash investment necessary on the part of the road electrifying that incident to the cost of the distribution and contact system required for the operation of its trains, and such buildings or modifications of buildings as will cover the shopping requirements of the new electrical equipment.

Commenting on Mr. Alfred W. Gibbs's valuable contribution in the discussion: His first point is well taken, and has been long recognized as one of the serious capital and operating expenditures where the problem of electrification is applied to deviating lines entering a city, and unless, as Mr. Gibbs points out, each one of these lines is electrified throughout the entire division, it is necessary to arrange local terminals to permit a change from steam to electrical operation. Such a condition, for example, is manifest in the study of the electrification of the various lines of the New York, New Haven and Hartford, the Boston and Albany, and the Boston and Maine Railroads entering the city of

Boston. As Mr. Gibbs advises, this is quite properly a charge to electrification, and a heavy one. On the other hand, large cities gridironed with their entering roads may have a density upon these tracks which will permit the terminal charges and still be within the economic limit of electrification. The tendency will be, of course, to electrify roads entering the cities upon which the greatest traffic obtains, and those having a lesser density will in turn be electrified as increased density justifies. The difficult financial situations in which the railroads of to-day find themselves are becoming better appreciated daily by the public, and it is safe to assume that, as density of traffic is the controlling factor in electrification, a city's people will see the justice of such a proposed procedure.

I fear my remark with regard to the third rail “having gone to seed" has been given more weight by Mr. Gibbs than was intended. Let me hasten, therefore, to withdraw any seeming attitude of criticism on my part with regard to the engineers responsible for the introduction of the direct-current system, particularly in the New York City terminals. But I confess that the adoption of the overhead system by the Pennsylvania Railroad Company for its Philadelphia electrification offered partial grounds for the suggestion that the third rail was past its flowering stage.

Mr. Gibbs's table showing the economy of steam locomotives with more liberal boiler capacity and with superheat is most interesting. The note that Mr. Gibbs adds, however, that “ These are the rates when running, added to which are certain standby losses at terminals," transforms "test" into practical conditions, and, as explained in my commentaries on Mr. Henderson's discussion, the improvement in thermal efficiency of central power stations can easily keep pace with any similar improvements in the steam locomotives. I feel sure that Mr. Gibbs will later find that my statement with regard to this matter is correct, and, if he does so, will take the first opportunity to withdraw his criticisms of my statement that "one pound of coal burned under the boiler of a central power plant will develop twice the drawbar power that the same amount of coal will produce when burned in a locomotive firebox." When I made the statement I had, of course, reference to the average conditions of practical operation.

As explained in the paragraph forward to my commentary

on the general discussion, the figures for depreciation upon the Cos Cob power station were not included, and for comparative purposes they should have been allowed. Mr. Gibbs has named a rate of 0.35 cent or 4 cents per k.w.h. as the amount that should apply, and it is to be noted that upon the basis of eleven per cent. upon the total investment made, covering interest, insurance, depreciation and taxes, the figure 0.29 cent, while less, does not differ greatly from that suggested by Mr. Gibbs.

Mr. Wheeler discusses most interestingly two vitally important points on the matter of electrification:

(1) The form of electric power which may be standardized for all classes of train service.

(2) The inheritance in the electric zone of " steamtrained" operators.

With regard to (1), the question of the location of the rectifier on or off the locomotive will always admit of local analysis for the most economic result. Mr. Wheeler has pointed out some very interesting possibilities in this direction.

With regard to (2), the matter of electrical administration, I have already expressed myself at such length on this matter that it will be only necessary to acknowldge with interest Mr. Wheeler's concordant sentiments.

Mr. W. A. Del Mar has asked some very pertinent and interesting questions, which I am very glad to answer. With regard to the expenditures made to reduce telephone disturbances, and of altering the right of way to conform with electrical requirements, these were included in the general figure of $15,000,000 mentioned in the paper. It is of interest, however, to note that past experience has indicated a proper method of laying out the transmission and distribution system whereby automatic compensation for telegraph and telephone disturbances can be secured for a very nominal amount. For example, in the case of the electrification of the four-track lines between Stamford and New Haven, the arrangement of transmission and distribution for the most economic traction result proves the most efficacious for reduction of the telegraph and telephone disturbances.

It is of interest to note that Mr. Del Mar is appalled at the development charges of the single-phase system, and, while I am at a loss to understand how a "fact" can be presented in an "ingenious way," as an example of a reduction of costs, our

strain insulators in 1907, when we first began the electrification, cost $63. To-day, with factors of safety three times both the electrical and mechanical values they had in their, original installation, the cost has been reduced to $7. A longer experience in the handling of work-train service and the administration of construction work upon those work trains are points along the curve representing forty per cent. in the reduction of construction, etc. With regard to the four questions Mr. Del Mar has asked, I am glad to answer them as follows:

(1) Q. Do these mileages include or exclude yard switching? A. The mileages given for both passenger and freight service do not include yard switching (yard switching is taken care of by yard switchers designed especially for that purpose).

I would advise, however, that the locomotive miles as shown on line 2 of the operating statistics, in both the case of the passenger and freight service, include the mileage of locomotives in trains and also the miles "run light" between engine-houses and stations. In figuring the passenger operating costs, however, the mileages of engines "run light" between stations are also included, thus giving the total passenger locomotive miles, and in figuring the operating costs on freight locomotives likewise the miles of light moves and the mileage of locomotives switching at way stations on the main line are also included. It would doubtless be of interest to Mr. Del Mar to know that we have added a third sheet of statistical information having reference only to yard switching mileage and yard switching costs, which sheet had not been inaugurated at the time the paper was written.

(2) Q. If they exclude yard switching, do not the operating costs appear unduly high? If they include yard

switching, how is it estimated?

A. The above answer to Question I doubtless serves as an answer to Mr. Del Mar's second question. While agreeing to his statement that the costs are high, I only hope that I have made the reason for this clear, especially in view of the typical record of the costs of maintaining the ten engines which at the time of the paper had received a full overhaul in the new shops. As previously explained, some of these engines had run 300,000 miles without undergoing any general repairs.

(3) Q. Are light locomotive mileages included in train miles?

A. The train miles did not include the mileage of loco

motives "run light," but have been recorded and are included in the total locomotive mileage in computing the unit cost per locomotive mile both in freight and passenger service.

(4) Q. The fixed charges given for Cos Cob power station appear to be very low. It would be interesting to know whether they include depreciation.

A. As explained in the previous part of this discussion, they did not include depreciation.

Mr. Del Mar's point with regard to the negligible error of meter registration on locomotives is very interesting, and, so far as we have been able to determine, the meters have been an accurate and valuable adjunct in the determination of the general distribution of power.

The Fire Hazard in Turbo-Generators. G. S. LAWLER. (National Fire Protective Association, vol. 8, No. 4.)-The chances of electrical generators of the older types being seriously injured by fire in the event of some part of the insulation failing are slight. Their freedom from fire damage is due principally to the comparatively low speeds, the accessibility of the combustible insulation, and the large mass of the machines per unit of capacity. In the case of generators of the turbo type this condition is reversed. Undoubtedly manufacturing companies have given the subject serious thought and have employed all means practical at the present time to reduce the fire hazard to a minimum, but there is still very much to be desired. Besides improving the non-combustible properties of the coverings, it would seem feasible to provide means for cutting off the ventilating system and thus diminish the supply of oxygen, or even to introduce carbon dioxide in the place of air in the cooling system.

The Theory of Cold Light. W. D. BANCROFT. (Transactions of the Illuminating Engineering Society, vol. x, No. 4.)-It is claimed that all chemical reactions tend to emit light, and that they all emit light if made to take place very rapidly. It is shown that the luminescence of salts flames is a chemiluminescence, and the method of determining the reaction is outlined for the specific case of cupric chloride. If a suitable chemical reaction can be made to take place sufficiently rapidly without any marked evolution of heat, cold light is obtained. The firefly has solved this problem, though the nature of the substance which oxidizes is not known. The chemist will some day solve it in another way. The Moore light is probably a case of chemiluminescence, but most commercial forms. of lighting depend on temperature radiation for their efficiency.