ever did cause a trouble of this kind. The coach foreman at Corning, however, disconnected the water-raising apparatus and plugged its connection to the auxiliary reservoir, under each of these cars, when they returned to Corning. A few days later the temperature again dropped below zero, and the same cars came in to him with two of their triples frozen so they wouldn't work, and one pair of hose stopped up with ice-and the waterlifts proved an alibi.

"There was so much water in the airbrake parts, that it does seem hard to believe it could all come from the pump on the engine some people don't-won't believe it yet; but right there is where it came from.

"You have read it-I have explained it, time and again-that the moisture in compressed air will not precipitate that is, settle let go and fall down-while the air is hot or very warm. Now the pump heats the air as it compresses it, and the bigger the pump and the faster it runs, the more free air taken up and compressed, and the more heat generated. Where does the heat come from? From friction of the pump pistons? Not on your life. It gets it from the air it pumps in. No heat in the air when it's below zero? Oh, yes it has; it has heat that it gives up until it liquifies at 312 degrees below zero, and still has heat that may be extracted and the liquid air made a solid.

"If the air is compressed slowly, lots of the heat will be absorbed back to the surrounding atmosphere before it leaves the main reservoir, but if you run the pump at a pretty quick speed the air is still warm as it passes back through your brake valve.

"That night, when 31's engine backed into the train shed at Corning, the weather was very cold, the sky clear and the stars shining so brightly it wouldn't seem there was a bit of moisture in the air; but every bit of exposed metal was white with a coat of frost-where did the moisture come from to make that frost? From the same place it came from to make the "sweat" drops on the outside of that ice-cold lemonade pitcher, the hot test, dryest day last summer-from the atmosphere; the air was filled with it.

"When the engine coupled to the train it was nearly leaving time, and you know you want to go out from that station with full brake power through the 'checkerboard,' but there was only a few minutes to get it, besides trying the brakes, and the air pump had to hustle.

"The nine-and-one-half-inch pump makes pressure pretty quick, and think of the amount of air she was throwing per minute, hot, into the main reservoir. The walls of the main drum warmed up, and so did the inside of the piping back through the engine and tender. By the time the air pressure had passed through the first car, it had become cool, and now the moisture began to deposit itself. There are a good many slow-downs and stops getting out through that city; the brakes were constantly used and recharged-with more moisture-but still getting a little warmth from rapid compression until finally, when once out of town the limit of 70 pounds train line pressure was maintained with the feedvalve cutting off the supply; not much more warm air coming back now, and the water in the train line and triples commenced to freeze; at the next stop they had trouble.

"You are glad that you understand it, but you are all further interested in the remedy for it? That's easy, but it is not so easy to enforce its application. The heat from compression can be expelled by placing a cooling coil between the pump and main reservoir; the cooled air pressure would drop its water in the main drum, where it would be drawn off; but it is out of the question to talk of getting that coil put in, so we will have to head the water off after it gets past the brake valve, and the tender drain-cup is the best place for the enginemen to go after it. While she is pumping-up in the train shed, jump down and bleed it, and at every stop getting out through town; it may save time later on. Better take the torch along, as the little bleedcock will freeze up.

"The next night after it all happened, I rode out of Corning on 31 with Tracyhe's a freight man-only extra on that engine and he and his fireman kept jumping down and taking a drag at that drain-cock at every little stop-without any suggestion from me, either. When they stopped at Lakeside for water, Tracy placed his brake valve on lap, jumped down and held his torch to the draincup till he could get her to let the water down; then he ran ahead and opened the train line stop-cock by the cylinders, letting a rush of air out of the pilot house. After doing that and closing the tender bleed-cock, he was up on the engine before the fireman had finished taking water; and Tracy's air didn't freeze up. The point to this is, that while, of course, the drum should be bled every trip, don't

think that it catches a great amount of the water, and don't think that a dry main reservoir indicates a dry train line. WILL W. WOOD.

Boiler Explosions.

"I hear," said Charley Smith, as he sat down on a waste box, "that one of the big 'liners' burned its crown sheet at Harris the other day, and the strange part of it is the boiler did not explode."

"Then you think, according to that remark, that when the water gets low in a boiler that there will be an explosion?" asked Andy Johnson, as he sat down on the bench where idlers in the Talcott roundhouse usually congregated.

"No, not exactly, but what I do mean is this: Every time there is a boiler explosion it is caused by low water. Why doesn't the rule work both ways, and every time there is low water have a boiler explosion?" replied Charley Smith. "Well, what causes boiler explosions in your opinion?" asked Andy Johnson.

"I don't know as my opinion is worth much, but it is that the material on the outside from some cause is too weak to stand the pressure from within," replied Smith.

"The first thing to do," said Johnson, "is to find the causes and remedy them and avoid an explosion, and not wait until one occurs and then charge it to low water. While at times this is the cause, as a rule it is absurd, as was the practice in the early history of steam boilers, to charge their explosion to some unknown and obscure agency, and it is far more unjust, for it often puts blame on enginemen when there should be none."

"You do not believe, then, that low water will cause a boiler explosion?" asked Smith.

"I did not say that, but I do not be lieve that all boiler explosions, or even 25 per cent. of them, are caused by low water, although all of them are charged to that cause. It is an easy explanation and shifts the blame where it can not successfully be combated. If a crown sheet was exposed and was very hot, and the water that was in the boiler should prime or lift and part of it go over the crown sheet there would be generated a pressure gas or steam, whichever you please to call it-that would probably cause an explosion, or if the crown sheet became so heated that it would lose its power of resistance, then the pressure

would force the sheet from its fastenings. It takes some little time for a sheet bare of water to reach this degree of heat. There are plenty of cases on record where locomotives have been fired up without water in the boiler and, the fact having been discovered, the fire has been knocked out and after the boiler and firebox had cooled down they were gone over by a boilermaker with a calking tool and the engine was placed in service again without discoverable injury and no loss, except the use of the engine and the boilermaker's time.

"If this holds good for an engine under these conditions, why does it not apply where there is low water, if the crown sheet is not superheated or the water in the boiler primes? Your feed water enters at the check and gradually rises, cooling the sheet by degrees as it does so. The cooling would probably cause the boiler to leak; it might even cause cracks if the iron was heated so hot as to destroy the life of the iron."

"To what then do you attribute the majority of boiler explosions if not to low water?" asked Smith.

"To weakness of material from various causes. Take the bracing of the crown sheet, for example. When the engine is about to be turned out of the shop it is subjected to a severe hydraulic boiler test to discover any weakness it may have in its makeup. This done, the shop has done its whole duty. There will be an occasional staybolt inspection and there it stops. The boiler may be out two years, but the crown bar braces will be left severely alone; some of these might be broken and so covered with lime and scale from the water as not to be seen, and may even sound solid to the blow of a hammer. The brace-pin may be worn or half broken, and unless it is taken out it could not be discovered, and the same may be said of defects in the braces. 'Out of sight, out of mind' applies too often to parts of a boiler that can not be seen.

"Thorough inspection of boilers, boiler braces and staybolts will do much to avoid explosions, and in these days of pooled engines care should be taken by all concerned to see that the water glass and gauge cocks are kept in good condition."

"If they are not, they speak for themselves," said Smith, "and give you warning that they are out of order."

"Gauge cocks will, but water glass cocks are deceiving and should be watched closely," replied Johnson.

"You seem to lay great stress on the

It is noticed occasionally, however, that a cylinder is worn oblong or oval. While this may be explained in some cases by the two cylinders being sprung out of line through drawing the pump lugs down to a badly-fitting bracket, yet the writer once observed another cause which it may be of interest to describe.

An 8-inch pump had been reported by the engineer as not supplying sufficient air and a test demonstrated that the air cylinder was in very bad shape. After obtaining 100 pounds pressure in the main reservoir, the pump was stopped and the air head was quickly removed. Then, by running the pump very slow a strong blow was noted past the piston, but all at one place. After marking this point on the piston and removing the latter it was found that the openings in both rings were very large and were at this point, one immediately above the other.

The cylinder was perfectly dry, the side of its wall and of the piston opposite from the location of the ends of the rings showed recent, heavy wear while the reverse did not, the removed cylinder head was well coated with the ground off particles of cast iron and on the worn side of the piston these had so filled in about one ring as to render it solid, and, with the wear, remove all indications of there ever having been a groove there.

As there was ample evidence that the piston could travel true so far as other causes were concerned, the only explanation of the wear all taking place on the side opposite from the ends of the rings is that the leakage at the latter point in passing the piston forced it heavily toward the other side of the cylinder. The excessive heating resulting from hard service and a bad order cylinder, with lack of lubrication furnished the other requisites for rapid wear.

F. B. FARMER.

Terminal Brake Test of Engine.

It has been said that it does not require a first-class engineer to do nice braking with brakes in good order. While not strictly true, yet it emphasizes the fact that good order brakes are the first requisite for good braking, As defects will develop they should be promptly located, reported and remedied. To accomplish the first, certain tests are necessary, and as recent inquiries indicate an awakened interest in this feature, the following terminal brake test of engine is given:

2.

Air Cylinder Test.

Raise the pressure to about 90 pounds and open the air cylinder oil cup. Then, stop the pump and, when the piston rod is seen to be at rest, see if any air is blowing out of the oil cup. If so, there is a leak from the main reservoir past the air valves or their cages into the cylinder.

3. If there is no such leakage, start the pump and, with the oil cup yet open and the same main reservoir pressure, regulate the speed between 25 and 30 single strokes or exhausts per minute. Now, on each down stroke hold a finger just above, but not on the oil cup, to note if any air blows out. There being no back leakage from the main reservoir, any such blow at the cup will be leakage past the air piston packing rings, which should be reported if it appears to exist for nearly one-fourth or more of the stroke.

4. In all of these tests it is important to have about 90 pounds air pressure, and, in the last, that the stroke be regu lated as instructed.

than the standard type, place the handle in full release and, when the governor is regulating the pressure, note how much, if Where with any, variation takes place. any engineer's valve the governor varies to exceed 3 pounds, report the fact.

Engineer's Valve.

10. With standard train pipe pressure make a 10-pound service reduction, return to lap and note whether any increase in train pipe pressure follows. If so, report the fact.

Driver and Tender Brakes.

11. With 70 pounds train pipe and auxiliary reservoir pressure, make a 20pound service reduction, place the handle on lap, go quickly to one driver brake cylinder and note carefully whether the If so, repiston rod moves back at all. port that driver brake leaks off, stating where the leak or leaks exist.

12. Test and report the same for the tender brake. The piston travel of this brake should be reported for adjustment, stating the amount existing, when over 8 inches or less than 5% inches.

13. Each of the driver brake pistons should travel about the same amount, and when that of the two combined exceeds 10 inches, report for adjustment. The shortest travel, with that of the two sides com

Dialogue No. 16-Nine and One

Half-Inch Air Pump.

Instructor.-I expect that you want to talk about the 91⁄2-inch air pump this morning?

Student.-Yes, sir, that is what I came up for. I understand that the principal advantage with this pump over the 8-inch is in its being of greater capacity, and I would ask how much greater capacity it has?

Instructor.-In answering your question I do not think I can do better than to give you a test which I understand was made to show the difference in the caThe steam pacity of these two pumps. was 140 pressure used in these tests pounds, and the main reservoir was 261⁄2 x34 inches outside measurement. About 84 cubic feet capacity. The 91⁄2-inch pump required 38 seconds to compress the air from 0 to 70 pounds and the 8-inch

From 20 pump required 68 seconds. pounds to 70 pounds the 91⁄2-inch pump required 27 seconds and the 8-inch pump 50 seconds. The air cylinder of the 92inch pump will contain about 708.82 cubic inches of free air, while the S-inch pump's air cylinder will contain only about 408.64 From this it will be seen cubic inches. that the 91⁄2-inch pump is about twice the capacity of an 8-inch pump with the same steam pressure.

Student.-What is the stroke of the 91⁄2-inch pump?

Instructor. The stroke is 10 inches, and the air and steam cylinders are both of the same diameter, 91⁄2 inches.

Student. I should think there would be some advantage in having the steam cylinder of larger diameter than the air cylinder, like the 8-inch pump.

Instructor.-I do not think that would be of any advantage today. When the 8inch pump was designed engines did not