the case of smoke and gases, that the action begins from behind the paint and not from in front by the disintegration of the paint.

EXPERIMENTAL BARS EXPOSED EIGHT MONTHS-LOWER ONE-HALF PAPER, UPPER ONE-HALF PAINT, 3 COATS.

DISCUSSION.

The President.

THE PRESIDENT.-I should like to call attention to one point in Mr. Barker's experiments from which I draw a great deal of comfort, namely, that if we may trust his experiments and the samples which he has exhibited to us, there can be little doubt that the point which has so often been brought out here is fairly well made, namely, if you can put on metal a coat which is impervious to moisture, you have got protection. I do not see how we can look at Mr. Barker's samples, and not be convinced that the key to the whole matter is in keeping moisture away from the metal, and unless our protective coating does this, it will be a failure.

There is another feature which is somewhat instructive, namely, how thick actually is a layer of paint. Using Mr. Thompson's figures and considering that there are 231 cubic inches in a gallon, and that this is spread over 172,800 sq. in., we find that the coat of paint is actually less than 0.002 in. thick. Of course, if we put on two or three coats, it would be two or three times this thickness, but on the other hand, quite a percentage of many paints as applied, is volatile matter, which passes off, so that even at the best, every protective coating is a very thin affair, and perhaps we are expecting too much from it.

I should like to remark also, on another point, namely, that the brush is an element in the problem. We recently had an experience which brings this out very well. We specified the mixing of the paint to be used in painting steel cars, and the car manufacturers said that the paint was unusable and could not be spread. On investigation we found that they were using practically whitewash brushes, and it is needless to say that we did not modify our specification; but simply asked that they use such a brush as a man with an ordinary wrist could use successfully. There is no doubt but that the brush is an element in painting. Only one point further, namely, we have been puzzling a good deal of late as to whether it is possible to apply paint with a brush

and not leave air bubbles. Many practical painters say that it The President. is not, and of course we are told that we must use more coats, in order to overcome this difficulty. It is needless to point out that the hope of success with successive coats is based on the supposition that two air bubbles will not come, one on top of the other. It is plain, I think, that we have a large field for study before use.

WHAT IS THE BEST METHOD OF PAINTING STEEL

CARS?

By FRANK P. CHEESMAN.

The Committee appointed by the Master Car Builders' Association reported at their 1904 Convention as follows:

"Your Committee submits the following suggestions for

PAINTING OF STEEL CARS

"1. For New Cars: (a) The steel should be thoroughly cleaned of all rust and furnace scale before the car is assembled. (b) All joints before assembling should be thoroughly coated with coal tar. (c) After car is assembled all grease should be thoroughly removed from the steel and same given a good coat of carbon or graphite paint on the outside and underneath, and the inside a heavy coat of crude petroleum, coal tar applied hot, or some similar substance. (d) The outside to be given a second coat of graphite or carbon paint, as may be desired.”

"2. For Old Cars: (a) All scale and rust should be removed wherever it appears on the car, by steel brushes or scrapers, and in the case of the inside of the car by any of the above methods or by the use of pneumatic hammers or mauls. (b) After all scale and rust have been removed the car should be thoroughly cleaned with steel scrapers or wire brushes and blown out with air, in order to present a clean surface for the paint. (c) The methods of painting recommended for new cars should be followed out in the case of old cars, after a clean surface is obtained."

The following report was made at the 1904 Convention of Master Car and Locomotive Painters by a committee especially appointed to draw up the best specifications for

PAINTING OF STEEL CARS.

"It is the sense of this Association that, in the construction and painting of steel cars, the following points are of vital importance to their preservation:

"First.-All flash or mill scale, rust, oil, grease and dirt should be entirely removed from all parts entering into the construction of cars before any paint is applied. We believe that this can be best accomplished by the use of the sand blast.

"Second. During construction, all overlapping joints, wherever metal is placed upon metal, should be thoroughly coated with a heavy mixture of moisture repelling paint.