the action of minor details. On the contrary, the effort will be to establish the economic performance of a number of typical locomotives when operating under a wide range of conditions.

No locomotive or type of locomotives will be acceptable, the value of which has not been proven by successful service on the road.

Locomotives to be acceptable must have weight and power which will make them comparable in these respects with the modern American machine. It is proposed to test no locomotive which has less than 2,000 feet of heating surface in its boiler, excepting that in case of locomotives having superheaters, the superheating surface may be regarded as heating surface, and in the case of locomotives having Serve tubes, credit for the surface of the ribbing will be allowed.

The gauge of the supporting wheels will be 4 feet 81⁄2 inches, or the same as the standard gauge of American railroads, and the gauge of the locomotive offered for test must be such as to run safely thereon.

It is planned to test 12 different locomotives, and it is hoped that a portion of this number can be of foreign design and construction. The time to be allowed to each locomotive will vary from 20 to 14 working days, the longer time being allowed those which are tested early in the season when both men and equipment will be new to the work.

.

to be preceded by one or more preliminary runs for the purpose of checking the valve setting, and of proving all accessory apparatus.

Each formal test will involve a run of approximately 100 miles, and throughout its duration the speed, load, steam pressure and other conditions of running will be maintained as nearly as possible, constant. The conditions represented by the several tests upon each locomotive will be so chosen that the results will fall into sets, and when so plotted will serve to disclose the performance of the locomotive under the full range of speed and cut-off for which it can be properly worked.

Methods to be Followed in Running a Test. In preparation for a test, the locomotive will be started and gradually brought to the conditions of running which are to prevail throughout the test. When these conditions have been secured the preliminary running of the locomotive will be continued until the rate of firing becomes uniform and until all portions of the locomotive have become warmed to their work. When these conditions have been secured two strokes of a bell will give a preparatory signal. Thirty seconds later a single stroke of the bell will mark the beginning of the test. Upon this stroke all water levels will be observed, the ash pan cleaned and all observations taken, and thereafter all water and fuel used will be taken from a weighed supply. Throughout the test all conditions of running will be maintained as nearly constant as possible, observations being taken on the stroke of the gong at ten-minute intervals. The duration of the test will vary from two to six hours, depending upon the rate of speed and load. The element of control in fixing the length of the heavy power test will be the amount of water evaporated, no test being ended until the evaporation equals thirty pounds for each square foot of heating surface. The lighter power tests may end after from four to six hours.

A test will be ended as it began. The fire which, throughout the test will have not changed greatly in its condition, will be brought as nearly as possible to the condition it had in the beginning, the ash pan will be cleaned, the water level in the boiler will be made to agree with that of the beginning of the test, and upon signal the final observations will be taken, and the use of water and fuel from a weighed supply will cease. As soon as practicable

THE GREAT LOOP ON THE COLORADO MIDLAND RAILWAY. SCENE NEAR HAGERMAN TUNNEL

after this the locomotive will be stopped, the front end cleaned, and the data of the test collected and made of record.

A test will be started not earlier than 8 o'clock on each day, and when the conditions are such as will permit them to be of short duration, two tests may be run on the same day.

To avoid chances for error all important observations will be taken in duplicate by the use of independent instruments and observers. For example, the feed water will be metered and afterward weighed, the weighings constituting the real record, and the readings of the meter the check record. The speed will be indicated by a Boyer or other speed indicator, and also by a counter, which will register the revolutions, the latter supplying the real record, and the former the check record. Pressures will be observed from dial gauges, and registered by a Bristol recording gauge, the observed pressures constituting the real record; the recorded pressures the check. A separate indicator will be used on each end of each cylinder.

The smoke discharge above the locomotive will be so arranged as to entrap all solid matter or "sparks" passing out of the top of the stack. A chemical analysis will be made of the coal employed for each test, and of the smokebox gases.

In the case of locomotives designed with special reference to the balance of reciprocating parts, and in the case of others, the performance of which may contrast with them, an effort will be made to study the motion (rocking, nosing, etc.) of the locomotive as a whole while running at speed, in the hope that a definite relation will be found between the motion of the locomotive and its condition of balance.

In the case, also, of certain locomotives which will be selected with reference to their type of boiler, an effort will be made to secure a record of the direction and activity of the water currents circulating within several portions of the boiler when the latter is delivering steam, and especially of the cooler currents discharged from the injectors.

There will be obtained for each test, by direct observation, the following facts:

Position of reverse lever.
Position of throttle.

Revolutions per minute.

Total revolutions.

Pounds of coal fired.
Pounds

of non-combustible material collected in ashpan.

3.

2. The performance of the boiler. The performance of the engines. By having a separate presentation of the engine and of the boiler performance it will be possible to trace the effect of each modification in design, whether in the boiler or engine; that is, changes in boiler performance resulting from changes in proportions or forms, will readily be traced, and changes in engine performance resulting from difference in design can be accredited to their proper cause. Moreover, it will be possible in the final analysis of results to interchange the boilers and engines of different locomotives, and to predict with certainty the general results which would have been obtained from a locomotive made up of any such combination. All of these facts will

work.

Publication of Results.-In order that the results may serve the largest purpose

2. Bulletins of Results.-These will be issued from time to time, and will constitute formal reports of the performance of individual locomotives. It is expected

The

+ See Note + to Table J.

1. Bulletins of Announcement.-These will constitute the communications of the Pennsylvania Railroad system and its advisory committee to the public, with reference to their plans and purposes. They will give information to those who may exhibit locomotives for test, and it is hoped that they will draw from those who are interested, such suggestions or

have been completed the results obtained, and such analysis of performance as may be based thereon, will be entirely set forth in a bulletin.

3. A formal publication to be issued at the conclusion of the whole work, presenting in such form as may be hereafter determined, the facts and conclusions developed by the whole study.