employed here as the effect of the oil from the piston will soon soften the rubber, spoil the seat and gum up the piston.

Where with no leak from above the piston it fails to close entirely with a long train, but can be made to do so by jarring as just mentioned, the fault is more likely to be in a very leaky piston ring and dirty cylinder. With such a train a small discharge will reduce the train pipe pressure very slow. The dirty cylinder renders the piston hard to move and the leakage down past the ring prevents getting enough difference to drive the piston valve to its seat. This is an unusual fault.

When with a long train a reduction of 10 pounds or any similar amount is followed by an increase in gauge pressure and which results in the reduction being less than intended, the fault is not,

as commonly thought, due to a faulty piston packing ring. This ring is not expected to prevent upward leakage or from the train pipe to chamber D. To guard against such the leather gasket 32 is extended above the piston, and near the circumference or outer edge of the latter is a flange or ridge which strikes on the gasket and should make an air tight joint. Therefore, where such leakage occurs the fault is an imperfect joint against the gasket and is always due to the latter where the flange on the piston has not been damaged. Because of this and the fact that swelling of the gasket does not interefere with the proper action of the brake valve the gasket should never be trimmed or otherwise altered so as to prevent the necessary joint being made. It is also due to this, the need of uniform thickness else where and accurate location of ports that only the especially prepared gaskets furnished by the manufacturer should be used. It seems hardly necessary to say

Brake Valve Handles Hard. A hard handling brake valve is due to unusual resistance at one or both of two points. One is the contact of the rotary valve with its seat and the other the joint of the rotary valve key, No. 12, Fig. 1, made by leather washer 13. For either of these a heavy lubricant, in the nature of a grease without much tendency to gum, is best. For the leather washer a mixture of grease and graphite is believed to give the best results, but the graphite is not recommended for the rotary valve as it would likely cause a gumming up of ports. It is important that the prepared leather washer as furnished by the brake manufacturer, be used, so as to insure a uniform and proper thickness; also, because of the lubricant worked into the leather. A thin washer will allow the metal surfaces to rub, while one too thick will force the rotary valve hard onto its seat, both causing the valve to handle hard.

An imperfectly seated rotary valve, even though it does not leak, will often cause the same result.

Another cause is carrying a very high excess pressure. The higher the main reservoir pressure, the harder is the rotary valve key driven against its gasket. The main reservoir pressure above the rotary is in part balanced by the train pipe pressure where the bottom of the rotary is Hence, the greater subjected to same.

the difference between these two pressures the harder will the rotary be held against its seat.

dan

Valve Does Not Mark Positions.-A gerous defect which is liable to be given little consideration is a failure to mark the intermediate positions so they can be distinctly felt. Such should be immedi ately repaired as otherwise the valve handle may be brought from release to lap position when it was intended to return it to running position; or, it may be accidentally moved from running position to lap by the engineer striking it with some portion of his body while moving about in the performance of his duties. Cases are on record where with passenger trains having little leakage a failure to supply air to the train pipe

was followed by the pressure reducing so slowly in the latter that no brakes applied, but leaving little to do braking with when wanted. There is good reason to believe that some unexplained "air brake failures" have been due to this cause.

Improved Rotary Valve.-Fig. 3 shows the face of the improved rotary valve. The older one is different in the arrangement of cavity c, c. With it the entire cavity is open at the face. The improvement consists in extending across the middle of the cavity a thin bridge so as to give more bearing and wearing surface. The three small ports and c, c all open into the large cavity so that operation is exactly the same as before and

the new rotary can therefore be fitted to the seat of an old valve. The three small ports in a row are to prevent equalizing, port g being closed at any point from full release to running position.

Another cavity is the long shallow one p, which in service and emergency positions connects the preliminary exhaust port with the atmosphere. There are three ports through the rotary; a, the supply port used in release; r, the warning port which blows in this position; and j, the running position or feed port. These three and cavity p have not been altered. The small end of the warning port r is at the top so it will be less liable to stop up.

(The brake valve will be concluded in the next number.)

that a reduction of twenty

pounds in the train pipe pressure will set the brake as hard as it can be set, I do not quite clearly see how this can be so; it seems to me that the more the reduction the harder the brake will go on," said a fireman to me a few days ago, while talking about air brakes.

Nearly every engineman, I believe, is inclined to think about as this fireman did until he begins to see clearly the plan upon which automatic air brakes operate; for it is not unreasonable to suppose, at first thought, that if a reduction in the train pipe pressure of a stated amount will apply the brake with a certain force, a greater, or additional, reduction ought to apply it with greater force, no matter if the reduction be made as great as thirty, or forty pounds. But it will not; so, therefore, let us review what we have already read about the air brake, and also study the sketches which are given with this explanation, in order to find out why it will not.

Fig. 1 represents an ordinary cylinder or reservoir which for our purpose we will consider is an auxiliary reservoir attached to the body of a car. If we were to attach a pressure gauge to it, and open the communication between the reservoir and the gauge, the hand on the latter woula register zero pressure, as

mean that there was no pressure whatever in the reservoir, for there is a pressure of one atmosphere, or 14.7 pounds, present in it about which the gauge does not tell us anything.

This pressure of one atmosphere, 14.7 pounds, plus whatever pressure the gauge may register, is what is termed absolute pressure, in order to distinguish it from gauge pressure alone, or the pressure that does effective work in pushing out the brake pistons.

Now we will commence, supposing, this reservoir connected to another of equal size, as shown in Fig. 2, and that all connections to this reservoir are tight, to pump air into it, that is, to fill it with what is termed compressed air, until the hand on the gauge registers 70 pounds.

We will then have in reservoir No. 1, Fig. 2, a total pressure of 70 pounds plus 14.7 pounds, equal to 84.7 pounds. The total pressure, 84.7 pounds, is the absolute pressure in the reservoir; the pressure 70 pounds, as shown on the air gauge, is the effective pressure, or what is termed the gauge pressure.

In the pipe which connects reservoir No. 1 with reservoir No. 2, there is a stop cock which controls the communication between the two reservoirs. If the stop cock is turned so that the communication is established between the two reservoirs the air in the first reservoir

can flow into the second, and it will be evident that when this occurs the pressure in the second reservoir will increase, and that in the first will diminish, until the pressure in the second reservoir is equal to the pressure in the first. When this happens the pressure in the two reservoirs is said to have equalized.

The question which now comes up is: How high is the pressure after equalizing in the two reservoirs? To answer this we will have to bring to our aid a little of the arithmetic which the editor has so kindly furnished us for the past year or so.

To begin we had in reservoir No. 1 (Fig. 2) before we opened the stop cock a pressure of 84.7 pounds absolute, or 70 pounds by the gauge, and in reservoir No. 2 we had a pressure of one atmosphere, 14.7 pounds. For the sake of convenience and ease in understanding the matter under consideration, we will suppose that there is no air whatever in the second reservoir, as shown in Fig. 3-nothing but zero pressure absolute in it-and that the air which it did contain has all been crowded into reservoir No. 1, so that the latter will contain a pressure of 99.4 pounds absolute.

Now with pressure of 99.4 pounds absolute in reservoir No. 1, and with nothing or zero pressure in reservoir No. 2, it is easy to see that when the stop cock is opened between these two reservoirs, they being equal in size and capacity, that when the air pressure equalizes between them each will contain one-half of the total pressure contained in reservoir No. 1 before the cock was opened.

To find out what that will be we divide 99.4 by 2, and it will give 49.7

The foregoing is an illustration of what takes place when communication is established between two vessels or reservoirs, one of which contains air, or any other perfect gas, under pressure, and the other has zero pressure absolute in it.

The truth of this statement has been verified many times by experiments, but was first established by the experiments of Boyle and Mariotte, who found the law which governs the variation of the pressure and the volume of perfect gases. and which is expressed as follows:

The pressure and the volume of any perfect gas, the temperature remaining constant, will vary inversely.

dimensions of the auxiliary and the brake cylinder are always maintained the same as those given in the example, and the pressure is just 70 pounds by the gauge, that we will get a little over 53 pounds pressure in the brake cylinder after full equalization, and this is about 17 pounds less than the auxiliary contained before communication was opened between it and the brake cylinder. Therefore, with everything in good shape, a reduction of 17 pounds in the train pipe from an initial pressure of 70 pounds by the gauge ought to set the brake in full.

We have already learned that the triple piston in the triple valve has train pipe pressure on one side and auxiliary reservoir pressure on the other side of it, and that these two pressures are equal when brakes are released and the train

by the stop cock in our illustrations in controlling the communication between the auxiliary and the brake cylinder.

With our auxiliary and train pipe charged to 70 pounds we commence to apply the brake by moving the handle of the brake valve to the service application position, and to allow the train pipe pressure to escape to the atmosphere, the triple valve moves to the right, and allows the auxiliary reservoir pressure to flow or expand, into the brake cylinder; but after we have reduced the train pipe pressure about 17 pounds from the initial pressure of 70 pounds, the air in the auxiliary ceases to flow into the brake cylinder, because as we have shown in our illustration, the pressures have equalized, and, therefore, it is of no further service to exhaust any more air from the train pipe in the hope of increasing the pressure in the brake cylinder or of setting the brake harder. It would be a waste of air.

But here you might say, "17 pounds is not 20 pounds, and we are instructed to reduce the pressure from 20 to 25 pounds in order to apply the brake in full service application. How about that?"

pipe and the auxiliary are fully charged, and we also know that when the train pipe pressure is reduced, the auxiliary pressure remaining greater, the triple is moved toward the right or toward the train pipe pressure, and that in so doing it closes the communication between the auxiliary and the train pipe, between the brake cylinder and the atmosphere, and opens communication between the auxiliary and the brake cylinder.

Here is where the triple valve performs a similar service to that performed

That point is easily explained. In the explanation and illustration that I have given you, I have assumed everything to operate without loss from leakage or clearance, but in practice there is always some loss from leakage, and from the small amount of clearance space in the brake cylinder to fill, and experience has shown that this loss is sufficient to lower the point of equalization of pressure in the brake cylinder to about 50 pounds, gauge pressure, when the piston travel is 8 inches, instead of 53.3 pounds as shown