came right together in a mass and started to spread out like water Mr. de Wyrall. blisters. You could run a knife under them and find quite a mass of corrosion. At the time we started I took exception to the method of painting, that is, to the material that was being used. I said that under these conditions we needed something else besides linseed oil as a vehicle, for after saponification took place which would be the natural result where vegetable or animal oil was used, owing to the alkaline nature of the condensation, the coating would become porous, and most of the condensation would go right through it, leaving the surface of the paint intact and doing its deadly work underneath. Subsequent events proved this theory to be correct. We then made several tests, I think about one hundred, and finally we eliminated, for the above reason, any paint that contained linseed oil as a vehicle. Among the different coatings, we tried some that would not saponify. We now have three, and, of course, we are going to see which will last the longest. They have been on over two years and are yet in good condition. One proof of the difference in the efficiency of the coatings can be seen when a change in the outside atmosphere takes place. On the columns coated with leads and oil no condensation will appear for from 12 to 24 hours, while on the non-saponifying coating the moisture can be seen running down in 2 or 3 hours. The first absorbs the moisture until the coating is completely saturated, before showing on the surface, while the other being impermeable sheds it at once. I do not decry the value of laboratory tests, for I have found them extremely useful in many cases, but in this specific instance the value was nil, for the coatings that have out-lasted all the others in a comparative service test were condemned by the chemist of the Rapid Transit Commission. But as the conditions in our Subway are totally different from those in the laboratory, and as the coatings are good to-day, it is from them that we shall choose which to use in repainting. Our finish coating at stations is white enamel and to be effective as a preservative coating it should always be elastic. We had great difficulty in getting the two materials to combine, owing to the difference in the coefficients of expansion of enamels and paints. This has been overcome, and we now have an ideal finish for our stations. At first we also had some difficulty owing to the work being rushed, one coat following another far too Mr. de Wyrall. quickly-sometimes on the same day-and the outer coats naturally cracked. But as we are now allowing ample time between successive coats for drying, the result is very satisfactory. Mr. Evans. Mr. de Wyrall. The conditions on the Subway are so different from any other, that a different material is required to meet them. While standard specifications will answer for a great many other products, in the case of preservative coatings, you can not expect one standard specification to fit all conditions. I made an interesting experiment in the Subway the other day, regarding the effect of electrolysis. I had a coating brought to me which was said to be an insulator. I painted a section of the contact rail two coats and failed to get an arc through it, all though the rail is charged with 700 volts. If this coating wilstand a time-test we will be able to minimize the disintegration of steel caused by electrolysis to a great degree. I hope to report the result of this at our next meeting. MR. S. M. EVANS.-Were the paints referred to as drying in ten minutes linseed-oil paints? MR. DE WYRALL.-No; these dried by evaporation. MR. TOCH.-With reference to the Subway, permission was granted me to make an investigation as to the cause of the oxidation in the Subway, and, on the May 31, 1905, my report was published. It took a long time for me and my two assistants to establish the particular nature of the oxide found. In my report I note a great many analyses of the Subway rust, which is a peculiar one, and one we never found anywhere else. The composition is Fe,O,2H2O+Fe ̧O,H2O-a very peculiar combination. We made over a hundred analyses to determine every possible condition, and I think the work was done very fairly, indeed. Strange to say, at the time I was doing this work I came here to Atlantic City for a few days' recreation, and in walking under the steel pier I noticed, at mean tide, the steel posts were badly rusted. I took off scales to in. thick and took them to my laboratory. I have a comparative table here showing that the Subway rust and that produced here are identical and not Fe,O,3H,O-the old formula for rust. The rust in the Subway was a yellowish brown streaked with black, just the same as the rust here on the steel pier. In making these comparative tests it was peculiar to find that when rust forms that way it sometimes skips a portion of the iron and forms another scale back of it. PROTECTION OF IRON AND STEEL STRUCTURES MEMORANDA OF ELEVEN YEARS' TESTS OF VARIOUS PAINTS. BY LOUIS H. BARKER. About eleven years ago experimental investigation was begun with numerous well-known and established iron paint-preservatives, in order to ascertain by actual exposure tests the best one to resist the destructive action on steel structures of sulphurous gases in the form of smoke combined with the moisture of steam, and since that time fifty or more paints and combinations have been tried. As will be seen by the accompanying table, the list included many kinds of asphaltum, rubber, graphite, carbon lead and iron paints, and though the results showed varying degrees of resistance, it is remarkable that even with three coats of paint not one was found that did not show rust in less than a year. Of course, it is to be understood that the exposures were made so as to subject the test bars to the severest action possible in order to obtain the quickest results. In making the first series of tests new steel plates 10 inches square were used. As, however, the adverse conditions we were trying to overcome related to rusty steel, which is more difficult to preserve than new steel, rusty plates were substituted in all tests thereafter. And to still further endeavor to meet the existing conditions new plates were hung up and exposed to the smoke fumes until they became covered with sulphur scale, the thought being that an oxide scale due to atmospheric exposure might give different results. This scale or rust formation on these new plates apparently varied not only in amount, but also in the time of its formation, supposedly due to different chemical composition. As this might again give some variations in the experimental results, in order that all paints should be on as like footing as possible, angle bars 11 feet long were made use of and, as before, hung in the smoke until rusted, then cleaned with wire brushes, each foot of the bar painted with a different paint and again hung up. The results, however, continued to be unsatisfactory. In examinations of the test bars from time to time it was seen that upon many of them the paint was intact, but with protruding points which upon being pricked were found to be small rust formations pushing up the paint from behind, clearly indicating that it was not the failure of the paints but the rust action on the inner surface that caused the damage. As no rust can form without the presence of moisture, and as all paints are pervious to moisture (as Dr. Dudley's careful investigations of the subject have proved) this led to the conclusion that it would be necessary in some way to tightly seal the surface. Many kinds of materials for doing this were tried, with as many different results, until three years ago it was decided that a cheap paraffine paper answered the purpose best of all, and since that time all experimentation has been along that line. The few test bars that have been brought along and exhibited indicate the results. Besides the experimental bars referred to, the paper covering has been tried in a small practical way against smoke action, and after two years and three months exposure an examination of very recent date shows the outer paint, the paper and the first or adhesive coat all intact and, in many places where paper was removed for examination the adhesive coat not yet dry and the surface of steel the same as when painted. With such satisfactory results from this paper-process in the smoke tests, it was concluded to make a large-scale application and severe test on a large number of eye beams supporting a floor over and within a few feet of salt water and upon which the rust was due not to smoke but to the almost continuous dampness and presence of sewer gases. This was done over a year ago, and up to this time indication of damage of no kind is apparent. The mode of application of the paper is as follows: After the rust is carefully cleaned off by means of stiff wire brushes, a certain kind of tacky paint is applied, the paper then covered over and tightly pressed upon the painted surface, the joints of the paper slightly lapping. As soon as the paper is in place, it is ready for the outside coat of paint. It will be observed that by this process, the first coat of paint, the paper and the coat of paint over the paper can be applied with one scaffolding, thereby greatly reducing the cost, especially in high and dangerous places. These experiments, extending over only three years, are of too short a duration to determine the value of paper as a protection for iron and steel, but they certainly bring out the fact, at least in a These tests *Protectus. + Protectus. *Protectus. + Protectus. *Toltz Asphalt. Toltz Asphalt. *Toltz Asphalt. †Toltz Asphalt. *Non Corrosive. †Non- *Non Corrosive. †Non Corrosive. *Red Lead. Onyx Structural. †Red Lead. Onyx Structural. *Tite Cote. and exposed when drated with two coats of the different paints, one exposed when dry. C These tests removed and cleaned with wire brushes. Half d This test wa NOTE |