UC-NRLF 
 
 31 fl?fl 
 
LIBRARY 
 
 OF THK 
 
 UNIVERSITY OF CALIFORNIA. 
 
 GIFT 
 
 Class 
 
 /* 
 
heory and Practice 
 
 OF 
 
 PAINTING ON METAL 
 
 BY 
 
 A. HL SABIN, M. S. 
 
 PUBLISHED BY 
 
 DWARD SMITH AND COMPANY 
 
 45 BROADWAY, 
 
Rusted steel plates; upper plate cleaned, lower showing "tubercles" of rust. 
 
 (Frontispiece.) 
 
THEORY AND PRACTICE 
 
 OF 
 
 PAINTING ON METAL 
 
 BY 
 
 A. H. SABIN, M. S. 
 
 PUBLISHED BY 
 
 EDWARD SMITH AND COMPANY 
 45 BROADWAY, N. Y. 
 
Copyright, 1905, 
 
 BY 
 EDWARD SMITH AND COMPANY 
 
THEORY AND PRACTICE OF PAINTING 
 
 ON METAL 
 
 i. 
 
 IN 1898 the writer put out a little book of specifications for 
 painting on structural metal, which has been received with con- 
 siderable favor; but the last edition (of two) is now out of print 
 and it has seemed desirable to reprint the specifications and 
 to add such remarks as further experience has suggested. The 
 specifications themselves are almost unchanged; but some work 
 has recently been done which exceeds in excellence and thorough- 
 ness anything heretofore described, and it seems well to give^a 
 somewhat detailed account of it, which will be done in proper 
 course. 
 
 Uncombined metallic iron is so rare, as a natural mineral, 
 that practically there is no such thing. It is found chiefly as an 
 oxide, though carbonates, silicates, and other salts abound and 
 in particular the sulphide, pyrite, is very widely distributed. It 
 may be said that the coloring-matter of most of the dark-colored 
 minerals and rocks is mainly iron. Metallic iron is obtained 
 from the ore by using the most powerful chemical agencies and 
 it constantly tends to go back into combination; when it does 
 this by natural means the product is termed rust. Usually rust 
 is hydrated oxide of iron, mixed with more or less carbonate, and 
 often with sulphate, sulphide, chloride, or other salts of iron. 
 The common cause of rust is the action of air containing moisture 
 and acids, especially carbonic acid; and to prevent it we must 
 exclude water and gases. This is very difficult, but has been 
 attempted in many ways. 
 
 5 
 
In the first place, by embedding the metal in cement. Bridge 
 floors are sometimes covered with melted asphaltum an inch 
 or more in thickness; a good result may thus be obtained, but 
 as the material is of a viscous nature and tends to flow slowly 
 off a surface not level and confined, it is of limited application. 
 
 Of far greater use is Portland 
 cement, or concrete made of it; 
 this is strongly alkaline, not 
 only with caustic lime, but the 
 action of this on the potassium 
 and sodium compounds origi- 
 nally present in the clay from 
 which the cement is made sets 
 free caustic potash and soda. 
 These strongly alkaline bodies 
 naturally prevent the access of 
 acids and thus prevent rust- 
 ing. The best Portland-cement 
 mortar is somewhat porous; 
 and concrete is liable to con- 
 tain voids, which increase its 
 permeability; cracks often oc- 
 cur, and it is known to be 
 exceedingly difficult to make 
 concrete even approximately 
 water-tight. Whether steel 
 which is to be embedded in 
 concrete should or should not 
 be painted is a question about 
 which there is no sort of agree- 
 ment. On one side it is said 
 that cement naturally forms a 
 strong bond with a metallic iron 
 surface, and that even if it is 
 a little rusty this rust will be 
 taken up by the cement. The 
 recent and apparently careful 
 tests by Breullie appear to show 
 that there \ is; considerable ad- 
 hesion between Portland-cement mortar and steel or iron surfaces, 
 and that this adhesion is determined by chemical action ; also that 
 the compound thus formed is soluble in water, that if it be removed 
 by percolation the bond is destroyed, and "that when it becomes 
 
 6 
 
 Stand-pipe at Racine, Wis., 135 ft. 
 high. Steel tank painted with 
 Durable Metal Coating. W. H. 
 Laing, Supt. 
 
necessary to rely upon this bond to preserve the integrity" of 
 any composite structure "which is exposed to the action of water, 
 it becomes necessary to make the concrete impervious to water." 
 (Eng. News, Vol. LI, pp. 561, 562.) 
 
 M. Breullie made also a series of experiments to determine 
 the strength of this bond existing between mortar composed of 
 .1 part cement to 2 parts sand and iron. His conclusion was that 
 after 30 days it amounted to 22 pounds per square inch of sur- 
 face. When we consider that this is only 10 per cent of the 
 strength of the concrete and a twentieth of i per cent of the 
 strength of the steel, it seems of very doubtful value; and the 
 almost universal practice of using twisted bars, or those rolled 
 in irregular shapes so as to make anchorages, which utilize the 
 rigidity of the cement, seems fully justified. Metal wliich is 
 embedded in concrete is, however, strongly held in place by the 
 pressure against it, caused by the shrinkage of the cement in the 
 act of setting, provided the concrete sets in air; if it sets under 
 water it expands. Metal which is painted will be gripped in 
 this way quite as well as that which is unpainted; and if the 
 action of percolating water is liable to cause rusting a coat of 
 paint ought to have a good effect. The whole matter has not 
 been sufficiently studied; but M. Breullie's experiments certainly 
 seem to show that the claims made for the existence of a chem- 
 ical "bond" of high efficiency between concrete and metal are 
 extravagant and not founded on measurements. 
 
 In any case only a small percentage of structural steel can 
 ever be embedded in concrete. It has been proposed to paint 
 iron with a cement wash; but this is brittle, and, so far as my 
 own observation has gone, does not adhere well. Newberry 
 says that in order to have it adhere well it should be kept moist 
 for a day or two, which is manifestly difficult. We come, then, 
 to the use of films of paint and varnish to prevent the action of 
 moisture and air. 
 
 Ordinary oil paint is composed of linseed-oil as the principal 
 liquid ingredient, with which is mixed some pigment; which 
 latter is a solid substance very finely powdered. The pigments 
 in principal use for the protection of metal are white and red 
 lead, the former more or less mixed with barytes and sulphate 
 of lime and sometimes with the carbonates of baryta and lime, the 
 latter with iron oxides, graphite ,'boneblack, and lampblack; with 
 these are mixed chrome yellow, chrome green, and ochre as tinting 
 materials. The lead pigments and lampblack are in their nature 
 chemical precipitates, and are extremely fine; but the others 
 
 7 
 
are for the most part pulverized by grinding; and it is a general 
 opinion that a given pigment is better the finer it is. Experiments 
 in regard to the amount of pigment required are not in accord 
 with each other; at one time it was commonly believed that 
 the greater the proportion by weight of pigment the better was 
 the paint ; now it is claimed that the finer the pigment the better ; 
 but the finer the pigment the smaller is the proportion which 
 can be mixed with : the oil, so far as weight is concerned. The 
 fact probably is, that the greater the number of particles of pig- 
 ment in a given measure of paint the better it is, or else that 
 
 the greater the aggregate surface of 
 these particles the better; the latter is 
 the more likely to prove correct, and 
 would make reasonable allowance for 
 the relatively coarse condition of graph- 
 ite, which cannot commercially be 
 ground to an extreme degree of fineness, 
 but which is in thin flakes, presenting 
 a large amount of surface. 
 
 The use of pigment in a paint is 
 threefold: it makes liquid paint thicker 
 than oil and thus causes a thicker film 
 to adhere to the surface; being com- 
 posed of solid particles, it hardens the 
 film and thus enables it to resist abra- 
 sion; and the particles of pigment fill 
 up many of the pores in the oil film 
 and so make it more impervious to air 
 and moisture. The use of the oil, on 
 the other hand, is to form a water-proof 
 coating as impervious as may be to gases 
 as well as liquids, and to act as a cement- 
 ing material or binder, to hold together 
 the particles of pigment and attach them 
 to the surface. The best oil for this pur- 
 pose is probably linseed w ich differs from almost all other oils in 
 the fact that when it is exposed to the air it is converted into 
 a dry, tough, leathery, or rubber-like substance known as linoxyn. 
 This change is caused by chemical action; the oil absorbs and 
 unites with oxygen, which it takes from the air, and so increases 
 in weight; it dries, but not in the way a watery solution does. 
 The film thus formed is not entirely impel vious ; it is quite insolu- 
 ble in water, dilute acids, and all ordinary solvents; but it is 
 
 8 
 
 Stand-pipe at Sedalia, Mo., 
 of the M. K. & T. R.R. 
 S. B. Fisher, Cons. Eng. 
 Painted with Durable 
 Metal Coating. 
 
somewhat porous. The supposition has been advanced that it 
 absorbs oxygen, combines with it to form carbonic acid, and 
 that the escape of this gas causes the formation of holes, or pores; 
 this seems unlikely, because if this action took place it would 
 be a true .combustion and would be accompanied by a loss of 
 weight; on the contrary, it increases in weight; the increase 
 being stated by Livache at 16 to 18 per cent and by Andes at 
 14 to 19.7 per cent. Further, if the pores in an oil film were blow- 
 holes it is difficult to see how the particles of pigment would 
 close them; the latter would be pushed out of the way; but 
 in fact nothing is better established than that the porosity of 
 the film is greatly diminished by the presence of pigment. The 
 cause, then, of the porosity of a paint film must be sought in 
 some other way. 
 
 Let us look at it in this way. A gallon measures 231. cubic 
 inches. Spread over 400 square feet of surface this would make 
 a layer .004 of an inch in thickness. Now, I have measured with 
 a micrometer-caliper a great many dry films, and two coats of 
 dry paint seldom measure more than .004 and the most I have 
 ever seen was .006; these were on test-plates, where care was 
 taken to lay on full heavy coats, and the proportion of 400 square 
 feet to the gallon (which is more than the average of commer- 
 cial work in painting on iron) could hardly have been exceeded. 
 Hence it appears that the film shrinks from 30 to 50 per cent 
 in bulk. This means also that the specific gravity of the oil 
 increases, which is the fact; I have never been able to oxidize 
 linseed-oil without its containing air-bubbles, but notwithstand- 
 ing this the oxidized oil is heavier than water; so it may be 
 regarded as certain that considerable shrinkage in volume takes 
 place. When an oil film dries it turns first to a gelatinous sub- 
 stance, or becomes tacky; this doubtless prevents its flowing; 
 and then as it shrinks, the contraction causes pores to form in 
 the film, which is drawn apart into a sort of network, differing 
 from an ordinary network in that the openings occupy less area 
 than the solid portions. The smallest of these pores probably 
 do not let water pass through, unless under pressure, because 
 of their capillary action; but it enters some of the larger ones. 
 If the oil is mixed with particles of pigment, which do not con- 
 tract, we would naturally expect these to stop up many of these 
 holes in the film, and that is certainly what takes place. This 
 explanation is simple and reasonable, agreeing with all the 
 facts. 
 
 Supposing this explanation of the lack of continuity of a 
 
 9 
 
paint film to be the true one, as I believe it to be, we may next 
 consider other means of making a less porous coating. 
 
 Films which are approximately water-proof may be made of 
 
 Stand-pipe at Lawrence, Mass. Steel tank painted with Durable 
 Metal Coating. 
 
 varnish. The most durable varnishes are those which have 
 linseed-oil as their elastic ingredient, with which is united a pro- 
 portion of resinous matter. Coatings of this nature have been 
 known from very remote times; the exterior wrappings of the 
 Egyptian mummies were varnished, and the wooden boxes, or 
 mummy-cases, in which they were laid, were thoroughly pro- 
 tected with thick coats of what undoubtedly is an oleo-resinous 
 varnish, which is still in good condition, twenty-five centuries 
 after its application; it may be seen in any of the large museums 
 of antiquities. No oil-paint has ever shown such durability. 
 It is no new thing to use such varnish as a preservative on iron. 
 In a manuscript of the date of 1520, in the library of San Marco 
 in Venice, is a recipe for varnish for iron armor consisting of 
 two pounds of linseed-oil to i pound of a varnish resin; another 
 author, Mathioli, in 1549 said that "with resin and linseed-oil is 
 prepared the liquid varnish which is used for giving lustre to 
 pictures and for varnishing iron"; and a Venetian author, Cane- 
 
 10 
 
parius, in 1619 said that with resin and linseed-oil is made a var- 
 nish which "adds splendor to iron and preserves it from rust." 
 Similar instances may be quoted down to the present time; this 
 use of varnish has been more common in Europe than in this 
 country; for example, the Eiffel tower was painted, when it 
 was built, -with a paint in which an oleo-resinous varnish, not 
 oil, was the liquid with which the pigment was mixed. But it 
 is not unknown here ; the author has made oleo-resinous varnishes 
 for painting hundreds of bridges and other important structures, 
 with the most successful results. It is, then, worth while to 
 inquire into the nature of these preparations. The resins from 
 which' varnish is made are of vegetable origin; but in most cases 
 
 Stand-pipe, Framingham Water Co., South Framingham, Mass. Height, 81 
 feet; diameter, 40 feet; capacity, 750,000 gallons. This stand-pipe is 
 coated with Durable Metal Coating. 
 
 they are fossil or semi -fossil resins, which after the fall and decay 
 of the trees by which they were produced, have laid buried in 
 the earth for a long time, in some cases hundreds or even thou- 
 
 1 1 
 
sands of years; during which time they have hardened and 
 have lost their perishable ingredients. To make varnish, a 
 quantity of some of these resins is melted in a suitable vessel; 
 to this is added a proper amount of hot linseed-oil, and the mix- 
 ture is then heated for some hours until the ingredients have 
 thoroughly combined. It is then thinned with spirit of tur- 
 pentine until its consistency is such that when cold it will be 
 suitable for use. When this liquid is spread out in a thin film 
 it has the quality of hardening, as linseed-oil does, by the absorp- 
 tion of oxygen from the air, and is converted into a coating which 
 is harder, smoother, and less porous than an oil film. It is natu- 
 ral to expect that it will be harder, because the resin which has 
 been added to the oil is harder than the dried oil ; but why should 
 it be smoother and less porous? 
 
 Nearly every one is familiar with the common use of rosin 
 for soldering; where, being melted, it acts as a flux, dissolving 
 and carrying away the solid particles or film of metallic oxides 
 which would otherwise prevent the contact of the solder with 
 the metal. All resins appear to be able to act as fluxes; and 
 when dissolved in oil they act in such a way as to make a com- 
 pound which, instead of being a stiff jelly which breaks up on 
 contraction into a network of holes, remains until it is dry (or 
 nearly so) a continuous, sticky, elastic film, nearly free from 
 porosity; and while the surface of an oil film is corrugated by 
 countless minute wrinkles, a varnish makes a smooth surface; 
 not only is this smoothness characteristic of the surface, but it 
 is indicative of the nature of the film throughout its whole thick- 
 ness; the film is continuous and of uniform structure. At least 
 it appears to be so, and that' is the general belief; but probably 
 at the very last some pores are formed, for such films, though 
 far less porous than oil films, do not form an absolutely perfect 
 protection, but are more satisfactory than any other coating 
 which dries in the ordinary way. 
 
 All good varnishes have this nature; but not all good var- 
 nishes are suitable for the protection of structural metal; because 
 if the proportion of resin is too large the varnish is so hard that 
 it is deficient in toughness, and cracks from sudden changes of 
 temperature. 
 
 It is impossible to have a surface too hard, if it is not brittle; 
 but it is absolutely necessary to have it tough. The larger the 
 proportion of resin the more brilliant is the varnish; and for 
 most purposes, that is, where varnish is regarded as an adorn- 
 ment as well as a protective material, it is desirable to have a 
 
 12 
 
high degree of brilliancy; moreover, there are many places, as 
 for example the outside of a railway coach, where the surface 
 is subjected to constant abrasion, where a considerable degree 
 of hardness must be had at any cost, even though a sacrifice 
 of elasticity and toughness result. So it is that all ordinary 
 varnishes are deficient in this respect; and a varnish particu- 
 larly designed for protective rather than decorative effect must 
 be made of somewhat different materials and in a somewhat 
 different way from those in ordinary use. Up to a certain point, 
 the larger the proportion of oil in a varnish the tougher and more 
 durable it is; but if more oil is put in than can properly com- 
 bine with the resinous matter it acts as a mixture, and then the 
 film begins to lose the characteristics of a varnish coating and to 
 resemble an oil film. This proportion cannot be definitely stated, as 
 it varies according to the character of the resin used (of which there 
 are many kinds) , and to some extent with the details of the process 
 of varnish-making, in which practical men differ considerably. 
 
 It is often said that as any true oleo-resinous varnish is, when 
 dry, harder than oxidized oil, it is therefore less elastic. No 
 doubt a varnish may be made with so great a proportion of resin 
 as to be lacking in elasticity, but it must be remembered that 
 flexibility and elasticity are essentially different things; and 
 while an oil film may be softer and more flexible, it may at the 
 same time be less elastic than a varnish film; and this is cer- 
 tainly true. Steel is probably less flexible than lead, but it is 
 more elastic, and wears better; and a baked varnish enamel 
 may be made which is far harder than any ordinary varnish film, 
 while it surpasses oil in toughness and is greatly superior to either 
 in freedom from porosity and in durability under all conditions. 
 So it is clear that hardness in a film is not necessarily an objec- 
 tion, but it is in part a desirable quality. It will be remembered 
 that the three objects for which we add pigment to oil are to 
 get a thicker, harder, and less porous film; varnish makes a film 
 which is harder and less porous; but how about thickness? To 
 this it may be said that pigment adds to the thickness of an oil 
 film by making the liquid paint thicker or more viscid than oil 
 alone; but varnish is in its nature more viscid than oil, hence 
 it may be laid on in a thick coating. It is reasonable, then, to 
 expect that a varnish not any varnish, but a special varnish 
 made of the most suitable materials and with proper knowledge 
 will be more durable than an oil-paint; it is smoother, tougher, 
 less liable to be affected by chemical action, and more homo- 
 geneous. Such we believe to be the fact. 
 
Practical experience supports this belief. The writer has 
 recently inspected an important bridge of considerable size 
 two plate-girder spans of 100 feet each which is subjected to 
 refrigerator drip, and over which fifty trains pass daily, which 
 has not had any repainting since it was built, more than ten 
 years ago; and was unable to find any rust on it, or any sen- 
 sible deterioration of the varnish, with two coats of which it 
 was originally coated. Another bridge which had a similar pro- 
 
 Steel stand-pipe, Winchester, Mass. Height, 40 feet; diameter, 40 feet. 
 This stand-pipe is coated with Durable Metal Coating. 
 
 tection has a draw 500 feet long; it is over sea-water, and tugs 
 pass under it continually; this was badly rusted in 1895 and 
 was then repainted ; in the opinion of the writer it needs repaint- 
 ing now, after nine years, but such is not the judgment of the 
 inspecting engineer, so it may be conceded that the treatment 
 has proved exceptionally satisfactory. It is not possible to get 
 as good results on an old and rusty structure as on a new one, 
 and the situation is a hard one as regards corrosion. In Chicago 
 there are a number of viaduct bridges about 1600 feet long, over 
 
railroad tracks, which were coated with this same varnish, one 
 coat at the shop and one coat after erection, in 1895 an( i 1896, 
 and have not yet needed any further care; and one of the large 
 train-sheds in that city has been efficiently protected in the same 
 way nearly as long. Scores of railway bridges can be seen which 
 have stood eight to ten years with nothing on them but the var- 
 nish with which they were originally coated; and in Boston may 
 be seen the great highway bridge connecting that city with Charles- 
 town, a structure protected in part with varnish and the remainder 
 with a varnish enamel paint, which has stood more than five 
 and in part more than six years; not so much as a rivet-head 
 has been repainted, and the surface is in fine condition. The 
 anchorages of the Williamsburg bridge, the greatest suspension 
 bridge in the world, which connects New York and Brooklyn, 
 were protected only with varnish; and the cables and steel floor 
 of the same bridge were treated with special compounds of the 
 same general nature, which will be described later. 
 
 It is not enough to use the best kind of a protective coating; 
 it must be properly applied. The first thing to be considered 
 is the preparation of the surface; it makes no difference whether 
 the coating is for protective or decorative effect, whether it is 
 a piano, a carriage, the outside or inside of a house, a train-shed, 
 or a railway bridge; whether it be wood, tin, iron, or steel; the 
 original surface must first be put in proper condition, or the 
 money spent on it will be likely to be wasted, at least in part. 
 There are people who think that paint forms an impervious coating, 
 and will therefore arrest any changes which may be going on 
 where it is applied; but none of those people are in the paint 
 business. The belief is absolutely universal among all those who 
 are making a living from the use of paint and varnish, that no 
 good results should be expected where the surface has not been 
 properly prepared; it may fairly be assumed, therefore, that 
 this belief is founded on knowledge, and the opposite belief on 
 ignorance, of the subject. It is the custom of those who apply 
 surfacing coatings to carriages, articles of furniture, and the like, 
 first, to have the article, if a wooden one, made as dry and clean 
 as possible; then to fill the pores of the wood with a suitable 
 material, and rub it to a smooth surface with sandpaper or other 
 abrasive agent, before the application of the paint or varnish. 
 In a much cheaper class of work, exterior house-painting, the 
 surface is made clean and dry, and the pores are filled with oil 
 or some mixture, before painting; and painting is done only in 
 dry and warm weather. Metallic objects are coated in a variety 
 
 15 
 
of ways; some receive a superficial layer of tin or zinc by im- 
 mersing them in a bath of the melted metal; some have a nickel 
 or silver layer electrically deposited; some are coated with a 
 vitreous enamel; and still others are dipped in a special sort of 
 varnish and then this is hardened by baking in an oven; this 
 latter process is called japanning. But in all these cases it is 
 regarded as absolutely necessary to clean off all the dirt, and 
 even to remove, by the action of acid or by the sand-blast, abso- 
 lutely all the oxide and scale, so as to show in all parts the color 
 of the pure metal, before anything else is done. If this precau- 
 tion is not observed, experience has shown that the coating will 
 not adhere properly; either it is easily scraped off, or if more 
 adherent, the difference of expansion causes it to flake off later. 
 If such is the case with so perfect a protective material as elec- 
 troplate or enamel, how can we expect it to be otherwise with 
 a much less completely impervious substance, such as paint? 
 There can be no doubt that the right way to paint structural 
 metal is to apply the paint, not to a superficial layer of dirt or 
 oxide, but to the pure, clean surface of the metal itself. This 
 can be done; the metal may be made perfectly clean, either by 
 the use of the sand-blast or by pickling with dilute acid, details 
 of which treatment will be found among the following specifica- 
 tions. 
 
 Sometimes it is not practicable to have the metal cleaned in 
 this thorough manner; it is therefore of interest to consider what 
 are the most and what the least objectionable among the things 
 which obscure its surface. Grease and mud are the worst, for 
 paint does not adhere to them at all; then comes a thick, recently 
 formed layer of ordinary rust. This is composed chiefly of the 
 hydrated oxide of iron; but it always contains, even though it 
 appears dry, moisture held in it as in a sponge, and with the 
 water carbonic acid; and these, imprisoned beneath a layer of 
 ^paint, soon assert their independence and regain their liberty. 
 Such a coating of rust is itself a cause of chemical action ; it is in 
 the most intimate contact with the unoxidized metal, to which 
 it gives up a part of its oxygen, quickly regaining from the external 
 atmosphere an equivalent amount; in this way it is a carrier of 
 oxygen between the air and the pure metal, and so it is that " rust 
 doth corrupt." There can be no safety until such rust is removed; 
 there should be no toleration of it. 
 
 Of quite another nature is the black, blue, or iridescent mill- 
 scale which is formed on the red-hot iron at the mill. This is 
 also an oxide, but is anhydrous, being either the magnetic or, 
 
 16 
 
more commonly, the anhydrous sesquioxide, equivalent to the 
 mineral hematite; and the principal objection to painting over 
 it is that its attachment to the underlying metal is not as per- 
 manent as could be desired. Its rate of expansion is somewhat 
 different, and if this causes it to come off the superimposed coat- 
 ing comes off with it. It is usual for it to consist of more than 
 one layer, and sometimes of several; in which case the outer 
 part easily scales off from the inner. This anhydrous oxide is 
 insoluble in acid; often the film next the metal is very thin, and 
 instead of peeling off on account of the unequal expansion, it 
 breaks up by microscopic cracks into pieces so small that they 
 are able to yield to small stresses caused by changes of tempera- 
 ture; such a surface is very satisfactory for painting, provided 
 it is taken in time, before rust has begun. The manner in which 
 this kind of scale is thrown off by chemical action is seen when 
 we put it in acid; the latter penetrates the minute cracks and 
 works its way under the bits of scale, which are thus loosened 
 and removed. 
 
 This suggests the treatment which should be given; all loose 
 scale should be removed by mechanical means and painting should 
 be done at once. Every one knows of the permanence of paint 
 marks put on at the mill while the metal is still hot; the paint 
 is melted into a sort of enamel and air and moisture never get a 
 chance to attack the metal. If the latter is partially cooled 
 by a spray of water, as is a common practice, and is then passed 
 through bending or straightening rolls, most of the detachable 
 scale is removed. In England it has been for many years a prac- 
 tice to coat the hot metal at this time, that is at a black heat, 
 with a thin film of oil, which bakes on by the residual heat in 
 the metal; this is done to steel rails and other metal work which 
 is to be sent on a long sea voyage, as to India or Australia, and 
 is found an efficient temporary protection against rust. It might 
 be an excellent first coat, preliminary to painting, on ordinary 
 structural metal; the film is thin, but hard and smooth; if desired, 
 it might perhaps be made of a sort of varnish, and be somewhat 
 thicker. 
 
 The floor of the driveways of the Williamsburg bridge is paved 
 with wooden blocks which rest on an underfloor composed of 
 steel channels, 12 inches wide and about 20 feet long, which are 
 laid side by side like a plank floor. These channels were sent 
 from the shop to an enamelling-plant ; when received there they 
 were covered with shop-grease and dirt, and were first put into 
 a tank containing a hot 10 per cent solution of caustic soda, to 
 
 17 
 
clean them. Next they were dipped in a tank of boiling water; 
 then in hot 10 per cent sulphuric acid; rinsed again in hot water, 
 then put in a hot 10 per cent solution of carbonate of soda, to 
 
 Laying the steel underfloor (coated with Sabin : Pipe Coating) on the 
 Williamsburg Bridge, N. Y. City. 
 
 remove all traces of acid, again rinsed in hot water, dried in an 
 oven, then dipped in a tank of hot varnish enamel, removed, 
 drained, put in an oven and baked at about 400 F. until the 
 coating was hard. They were then taken out and allowed to 
 cool, and were'^shipped to the bridge. When laid in place there 
 was not a scratch to be seen on more than one piece in fifty; and 
 after they were in place it was two or three weeks before they 
 were protected by the block pavement, during which time the 
 force of employees, numbering six or eight hundred, walked to 
 and fro on them, with wheelbarrows, casks of bolts, and the like; 
 and at the end of this time the coating showed practically no 
 injury. This was observed not only by the bridge engineers, 
 but by visiting engineers from all parts, and was greatly admired. 
 The sidewalks also have a steel underfloor, which was coated 
 in the same way. This was probably the first structural metal 
 (except pipe) which was ever enamelled, and, with the exception 
 of the metal in the anchorages of the old Brooklyn bridge, is 
 
 18 
 
believed to be the first structural metal to be cleaned by pickling 
 in the United States. Structural work has been sand-blasted 
 here, and has been pickled in Europe for many years. This 
 work I regard as the most thorough and perfect in the way of 
 protection agains.t corrosion which has ever been done. 
 
 Enamelling (by which is here meant the use of a varnish 
 enamel) has been applied to water-pipes for about ten years. 
 In 1893-4 a 38-inch steel-riveted pipe-line, about 28 miles long, 
 was laid to supply the city of Rochester, N. Y., with water. 
 Half of this pipe, about 14 miles, was protected by dipping the 
 sections, which were 28 feet long and weighed up to 2\ tons, in 
 a varnish enamel, designed for the purpose by- the writer, and 
 were then baked. These pipes were not pickled; but care was 
 taken that the steel plates should not be rusted, and in rolling 
 such plates into form they are put under great strain and all 
 the scale which can be got off is loosened; they were then care- 
 fully cleaned before coating; the enamel was thin, probably 
 not over .003 of an inch in thickness; but the coated pipes stood 
 
 Laying steel underfloor (coated with Sabin Pipe Coating) on Williamsburg 
 
 Bridge, N. Y. City. 
 
 the subsequent handling well, and have now been in service ten 
 years. Regular and frequent inspection has shown the coating 
 to be practically as good as when first applied; and in particu- 
 
 19 
 
lar it has not afforded a foothold for the growth of aquatic organ- 
 isms, a source of much trouble in ordinary water-mains. The 
 judgment of the engineer, Mr. Kuichling, who took the respon- 
 
 Steel tmderfloor (coated with Sabin Pipe Coating) ready to receive wood- 
 block pavement, Williamsburg Bridge, N. Y. City. 
 
 sibility of applying this coating, has thus been fully vindicated. 
 The following year the pipe-line at Allegheny, Pa., which is nearly 
 10 miles long and is 5 feet in diameter, was similarly coated; 
 these pipe sections were of i-inch plate and weighed 5 tons each; 
 the oven had a capacity of 8 sections, or 40 tons, and as much 
 as 64 of these sections, or 320 tons, was the output of a single 
 day of 24 hours. The same year about 4^ miles of 40-inch pipe 
 was similarly treated for Cambridge, Mass.; all this work has 
 been satisfactory. 
 
 Since that time the coating has been adopted for salt-water 
 pipes (fire-mains and flushing-mains) by the U. S. Navy Depart- 
 ment, and is now specified on all work of this kind on ships of 
 war. By its use the durability of these pipes has been increased 
 from fivefold to ten or twentyfold; and plants for its applica- 
 tion have been erected in nearly all the navy yards of this country. 
 Particulars regarding the construction and operation of these 
 plants will be furnished on application by the writer. 
 
 20 
 
In 1902 the company with which the author is connected did 
 some novel and interesting work in making a coating for the 
 wire suspension cables for the Williamsburg bridge. In the past 
 it has been customary to protect such cables by covering them 
 with white-lead putty and winding them with wire; this wire 
 is continuously wound from the towers to the middle of the cable, 
 the winding being as close as possible; this protects the cable 
 from abrasion and makes a very tight covering. But it is well 
 known that putty gets hard and brittle after years of age, and 
 may very likely crack from vibration and changes of tempera- 
 ture ; so it is the practice to protect the cables by frequent paint- 
 ing. It is not practicable to cut through the wound-on coating 
 of wire to inspect the interior of the cable, and the engineers 
 of the new bridge decided to use a new method of coating designed 
 in part by us. 
 
 Referring back to what has been said about the drying of 
 linseed-oil, and the use of resinous matters as a flux to avoid its 
 gelatinization, it may now be added that while it has for hundreds 
 
 Applying Edward Smith & Co.'s materials on the saddles of the cables of 
 the Williamsburg Bridge. 
 
 of years been known that oil can, by blowing air through it, or 
 by exposing it in films to the air, be converted into linoxyn, or 
 " oil-rubber," as it is commercially called, this oxidized oil (which 
 
 21 
 
is commonly made by allowing the oil to trickle down over sheets 
 of cloth, forming a coating which constantly increases in thick- 
 ness until it is thick enough to be mechanically removed) is of 
 
 Applying Edw. Smith & Co.'s materials on the cables of the Williamsburg 
 
 Bridge, N. Y. City. 
 
 value only for very few and limited uses; it is insoluble in all 
 solvents, and is not fusible by heat; if heated it chars but does 
 not melt; it is therefore a very refractory material to work with. 
 But many years ago. the writer made a study of the effects of 
 resinous fluxes on this material, especially to get such a com- 
 pound as would retain as nearly as possible the good qualities 
 elasticity and resistance to chemical action of the linoxyn. Such 
 a compound was made, which could be fused by a moderate 
 degree of heat, yet was quite unaffected by natural changes of 
 temperature; and in the liquid made by melting this, cotton 
 canvas was dipped and saturated, the excess of coating removed 
 as it was drawn from the dip, and when this coated canvas had 
 cooled it was perfectly air and water proof, yet was flexible; 
 and although it was so free from tackiness that it could be rolled 
 up on a spool, yet one layer could be made to adhere perfectly 
 to another by going over it with a hot iron. This coated canvas, 
 in strips about 10 inches wide, was wrapped around the cable, 
 lapping it enough to make in all three thicknesses; and a man 
 with a hot iron (such as is used in asphalt paving) followed the 
 
 22 
 
man who served on the canvas and, as it were, welded the whole 
 together into one impenetrable covering. 
 
 This may be cut open anywhere at any time for examination, 
 and repaired with more of the same material, as perfectly as 
 the original. The coated cable was finally covered with a sec- 
 tional tubular steel shield, to which the suspensory ropes are 
 attached. 
 
 The compound in question had been known to the writer 
 for several years, and its value and permanence are beyond doubt; 
 but the use of canvas coated in this way was new. It would 
 be impracticable to use an ordinary paint or varnish in this way, 
 because these require the accession of air to oxidize them, which 
 could not be provided; the compound used, being already oxi- 
 dized, is in a perfect and mature state as soon as it has been applied. 
 Canvas coated in this way may be made to adhere strongly to 
 steel or other metallic surfaces, and no doubt will, if it becomes 
 known, find many uses as a protective coating, as different from 
 paint on the one hand as it is from cement on the other. Not 
 
 Applying Edw. Smith & Co.'s coating material on cables of Williamsburg 
 
 Bridge, N. Y. City. 
 
 only is it ready for immediate exposure as soon as it is put on, 
 but it has the strength and resistance to abrasion of canvas, 
 which has long been used to cover the decks of yachts, steam- 
 
 23 
 
boats, and other vessels, as well as the roofs of street-cars; arid 
 the cotton fibre is not subject to decay if kept from the weather, 
 which condition is .exactly fulfilled by saturating it with this 
 
 Using a hot iron in cementing the cable coating (design and material from 
 Edw. Smith & Co.) on the Williamsburg Bridge, N. Y. City. 
 
 compound. It is not anticipated that such a coating will take 
 the place of paint, but may serve for uses for which we have as 
 yet had no satisfactory material. The introduction of a new T 
 kind of coating of merit and of practical and industrial avail- 
 ability is of sufficient importance to deserve more than a pass- 
 ing notice. The most serious defect of paint and varnish is the 
 thinness of the film and its consequent inability to resist abrasion ; 
 but a heavy cotton duck, weighing before coating three-quarters 
 of a pound to a square yard and upwards, is not open to that 
 objection. 
 
 Cleaning bridgework preparatory to repainting it is a matter 
 of great difficulty; but it is of the first importance, as all engineers 
 and all paint manufacturers agree that if it is not well done the 
 application of paint is of little use. The most thorough way 
 is by the sand-blast; this can be accomplished at a cost, probably, 
 of about two cents per square foot. The most important paper 
 on this subject is one by Mr. Lilly, of Columbus, Ohio, in the Trans- 
 actions of the American Society of Civil Engineers, Volume 50, 
 1903; but work in this line is being done in many places. It is 
 attended by the inconvenience of dust, which may in consider- 
 
 24 
 
able part, but not wholly, be obviated by screens of cloth. In 
 isolated places this is of little account. If for any reason this 
 is not thought practicable, it is necessary to thoroughly scrape 
 and then wire-brush the surface. Some bridge gangs are equipped 
 with complete sets of scrapers, some of which are wide, some 
 narrow; most are straight, but some are hooked like a hoe, to 
 be operated by pulling instead of pushing; they should all be 
 long enough to be firmly grasped by both hands, and a hammer 
 and chisel are sometimes useful auxiliaries. In 1903 the Chicago & 
 Northwestern Railroad cleaned their train -shed in Chicago by 
 scraping, after which a considerable part of the surface was washed 
 with acid, which was subsequently removed by alkali; then 
 washed, dried, and brushed before painting. In one of the largest 
 chemical works in the country iron is pickled with rather strong 
 sulphuric acid (the stronger it is, up to say 30 per cent, the more 
 rapidly it acts), then this is immediately neutralized with hot 
 milk of lime, no intermediate washing being done; the milk of 
 lime, made by slaking lime in water, is cheap material, and may 
 
 Applying Edw. Smith & Co.'s coating material on cables of Williamsburg 
 
 Bridge, N. Y. City. 
 
 be either rinsed off with water or allowed to dry on and then 
 removed with a brush. This is excellent practice. Iron treated 
 in this way is not so extremely prone to rust as is sand-blasted 
 
 25 
 
metal, but of course should be painted with the very least 
 possible delay. The wire brush is certainly a most useful and 
 efficient implement for getting dirt off metal; but it should not 
 be depended on for the removal of scale or thick and strongly 
 adherent rust. In our own factory we have a large rotary wire 
 brush driven by power, which is naturally more efficient than a 
 hand-brush ; but even this will not remove anything which adheres 
 very persistently. Thorough scraping should precede brushing. 
 If it is certain that the old paint which was on the bridge was 
 of good quality, which can be easily told at the time of repaint- 
 ing, and if it is in part in good condition, as is sometimes the 
 case, it is not absolutely necessary to remove it; and in fact I 
 think bridges should not be left without painting repairs until 
 the old paint is destroyed. A better way is to keep watch of the 
 structure and when any considerable break appears in the paint 
 let that place be cleaned and repainted. An engineer would 
 not wait until all the bolts and rivets rusted out of a bridge before 
 he began to put new ones in, and I see no reason why all repairs 
 should not be made in the same manner, 
 
 26 
 
Williamsburg Bridge, N. Y. City. Cables, floor, and anchorage metal work 
 protected with Edward] Smith & Co.'s coatings. L. L. Buck, Ch. Eng. 
 O. F. Nichols, Prin. Asst. Eng. 
 
 Charlestown Bridge, Boston. Wm. Jackson, City Engr. The girders were 
 painted with Durable Metal Coating, the rest of the bridge with Pelagic] 3 
 Enamel, made by Edward Smith & Co., N. Y. 
 
 28 
 
II. 
 
 SPECIFICATIONS FOR BEST WORK. 
 
 I. All iron or steel, after being shaped, punched, bored, and 
 otherwise made ready for assembling, must be cleaned by the 
 sand-blast or other equally efficient means, so as to show, in all 
 parts, the grayish-white color of the metal. 
 
 The cleaning of structural steel work by the use of the sand- 
 blast is probably the simplest and most satisfactory way to have 
 it done. The great objection to this, as to all such work, is the 
 cost. Since there are practically no plants for doing such work 
 in bridge , shops in fact there are no plants for cleaning steel 
 by any process any engineer who decides to call for thorough 
 cleaning must expect to pay an additional sum for such work, 
 which, if the work be of small extent, will be disproportionately 
 great in comparison with the work actually done; but if the 
 work be of great extent the incidental cost of the plant will make 
 but a small item in the total cost. Plants have been installed 
 since the publication of the first edition of these specifications by 
 several of the railroad companies for cleaning bridges in place, 
 by some municipal engineers, and by some manufacturing works 
 where structural metal is treated. It has been objected that this 
 is a patented process; it is therefore in order to observe that 
 the original patent, under which work of great extent was carried 
 on most successfully for fifteen years; was issued to Gen. Benj. G. 
 Tilghman, of Philadelphia, on Oct. 18, 1870, being No. 108,408, 
 and has therefore expired long ago. 
 
 Among the most important claims granted by that patent 
 were the following: 
 
 " i. The cutting, boring, dressing, engraving, and pulver- 
 izing of stone, metal, glass, pottery, wood, and other hard or 
 solid substances by sand used as a projectile when the requisite 
 velocity has been imparted to it by any suitable means. 
 
 "2. The artificial combination of a jet or current of steam, 
 
 29 
 
air, water, or other suitable gaseous or liquid medium with a 
 stream of sand, as a means of giving velocity to the sand when 
 the same is used as a projectile as a means of cutting, boring, 
 dressing," etc., etc. 
 
 "7. When a jet or current of steam, air, water, or any other 
 suitable gaseous or liquid medium is employed to give velocity 
 to sand used as a projectile, as a means of cutting, boring, dress- 
 ing," etc.; " the use of the following devices for introducing the 
 sand into the jet of steam, air, water, etc.: First, the suction 
 produced by the jet of steam, air, water, etc. Second, a strong, 
 close vessel, or sand-box, into which the pressure of the steam, 
 air, water, etc., is introduced, and through which, when desired, 
 a current of it may be made to pass." 
 
 It is obvious from the foregoing that there is no existing patent 
 on the process ; and while there is some patented apparatus which 
 is preferred by some of the people who use the process, this is 
 equally true of a very large proportion of all machinery in use. 
 The process of cleaning with the sand-blast is essentially as follows : 
 Air at a pressure of 20 to 25 pounds per square inch is furnished 
 by any suitable air compressor. If we assume that we will use 
 a discharging nozzle 9-16 inch internal diameter when new, 
 each such nozzle will require 120 cubic feet of air per minute, 
 measured at atmospheric pressure compressed to show a pressure 
 of 15 pounds per square inch at the nozzle. This is, how- 
 ever, to be regarded as a minimum, for it is advisable to 
 use a somewhat higher pressure, say 20 pounds; and tKe 
 nozzle rapidly wears away until it reaches a diameter of J 
 inch, at which it will discharge nearly twice as much as when 
 new, so that in practice it is well to provide an air-compressor 
 handling 240 cubic feet of air per minute and compressing the 
 same to 20 pounds per square inch. For four of these nozzles, 
 each connected with a No. 6 Ward & Nash apparatus, an Inger- 
 soll compressor of 24-inch stroke, 2 2 -inch steam, and 22-J-inch 
 air-cylinder was used; and such a plant ought to be supplied 
 with steam from a 150-!!. P. boiler. These latter figures are, of 
 course, rough; but from the number of cubic feet of air used 
 and the pressure any one can compute the rest of the plant. It 
 has been found that for the removal of hard, tough scale, a pres- 
 sure of 35 pounds per square inch is necessary. 
 
 Into this current of air dry sand is introduced at the rate of 
 about 10 cubic feet of sand per hour for each such nozzle, or 
 i cubic foot of sand to 1000 cubic feet of air. The sand must 
 be artificially dried. Some operators use coarsely powdered quartz ; 
 
 30 
 
this latter can be used five times in succession; and in general 
 the sand may be used until it is broken up into a powder too 
 fine for use. In the plants which the writer has inspected the 
 sand and air are carried to the nozzle through a heavy rubber 
 hose about 2^ inches in diameter. This is not worn away by the 
 current as a metal, pipe would be; but it is necessary that the 
 air should not be hot, as this would rapidly injure the hose. The 
 nozzles are short pieces of chilled-iron pipe, and have to be 
 renewed at frequent intervals. From data furnished me by 
 Naval Constructor Bowles I find that the cost of cleaning the 
 bottom of a ship in dry dock amounted to about 4 cents per square 
 foot; but this was done with an experimental plant, and the 
 method of drying the sand, which was used only once, was costly, 
 and the cost would certainly have been reduced to 3 cents per 
 square foot if a permanent plant had been in use. Since the instal- 
 lation of a permanent plant no work has been done of sufficient 
 magnitude to give figures. This was an exceedingly rusty sur- 
 face; but with this same experimental plant the mill scale was 
 removed from 3,55 square feet of surface of steel plates at a 
 cost of $17.60, or about \ cent per square foot, and at the rate 
 of 4} square feet per minute per nozzle. 
 
 Work recently done indicates that bridgework may be cleaned 
 in the field by this process for 2 cents per square foot. 
 
 It may be well to add that in all the work referred to, which 
 was practically field^ work, being"] carried on out-of-doors and 
 with a somewhat portable plant, the labor amounted to one 
 man to hold each nozzle, one man to attend to each two sand- 
 Doxes, and one man to clean up and carry sand for each four 
 nozzles. The supply of compressed air is an expense of a differ- 
 ent sort, as is also in field work the matter of staging, etc.; but 
 all are included in the prices given. It seems reasonable to sup- 
 pose that where many pieces of metal of the same general char- 
 acter are to be treated in a shop fitted up for the purpose, con- 
 trivances may be introduced which will do away with a con- 
 siderable part of the labor. 
 
 Iron and steel may also be cleaned by picklinsr in acid and the 
 subsequent removal of the latter. This may be done in the fol- 
 lowing manner : The pieces of metal which have been made ready 
 for assembling are immersed in hot dilute sulphuric acid having 
 a strength of 25 to 28 per cent. Some use acid of 20 per cent. 
 It is kept in this until the whole surface is free from rust and 
 scale; this will take from 6 to 12 minutes. If the pieces of metal 
 are somewhat rusty, so that rust has' started underneath the 
 
 31 
 
The steel framework of the Broadway Chambers Building, N. Y., was 
 
 painted with Ebonite Varnish. 
 
 32 
 
scale, the shorter time will be found sufficient; but if it con- 
 sists of plates covered with closely adherent blue or iridescent 
 rolled scale, the longer time will be necessary, since this scale 
 is itself insoluble in acid and is removed by the latter penetrat- 
 ing the innumerable minute cracks in the scale and attacking 
 the iron underneath, thus mechanically throwing off the scale. 
 If, on the other hand, the iron is uniformly rusty, this coating 
 of hydrated oxide readily dissolves in acid; and in fact a weaker 
 acid of 10 per cent to 12 per cent might be used, although the 
 stronger acid is quite safe, but will require a shorter time. It 
 has been suggested that it is desirable to previously clean the 
 metal with caustic alkali from all grease, etc.; but if acid of the 
 above strength is used^and kept as hot as possible this will not 
 be necessary. As soon as the acid has reached the iron in all 
 parts of the surface the metal is taken out and washed by jets 
 of water discharged against it under high pressure, not less than 
 100 pounds per square inch, and much better if double that. 
 In this way the acid may be thoroughly removed. 
 
 In Germany it is said to be customary to use acid of 9 or 10 
 per cent cold, and the metal is left in it five hours. This makes 
 a much larger plant necessary and has no advantages. 
 
 If it is attempted to remove the acid by soaking the metal 
 in still water the following difficulty is encountered: the iron 
 becomes immediately coated with a gummy or colloidal substance 
 very difficult to remove. What this is, is not known to the writer; 
 but it is well knownjthat there are a number of insoluble or diffi- 
 cultly soluble compounds of iron with sulphuric acid, and it is 
 probable that some of these are precipitated on the surface of 
 the iron when water removes the excess of acid; but if a jet of 
 water is used the mechanical effect .is to remove the adherent 
 ferrous sulphate at the same instant, leaving a clean metallic 
 surface. It is also possible that if the acid contains arsenic , 
 as is the case with much of the acid made from pyrites, this may 
 also be precipitated on the surface; in fact it is sure to be, and 
 acid free from arsenic should always be used for this purpose, 
 and as a matter of practice it is insisted on by many. 
 
 Very successful work is now done by (i) dipping the metal 
 in 10 per cent solution of caustic soda, boiling hot; (2) in boiling 
 water; (3) in hot 10 per cent, sulphuric acid; (4) in boiling 
 water; (5) in hot 10 per cent carbonate of soda; (6) in boiling 
 water; then put in the oven to dry. The process is economical 
 and the results are all that could be desired. 
 
 It is often difficult and sometimes impracticable to pickle 
 
 33 
 
steel high in carbon, and cast iron containing graphitic carbon, 
 on account of the deposit of a film of carbon, like stove-blacking, 
 on the surface. Muriatic (chlorhydric) acid has been used instead 
 
 One of a group of four bridges on the Boston & Albany R.R., all painted 
 with Durable Metal Coating. W. Shepard, Ch. Eng. 
 
 of sulphuric; but it is not well suited for the purpose, being much 
 more expensive and difficult to remove. It also forms a gummy 
 coating on the iron, worse than that with sulphuric; and in the 
 subsequent alkaline treatment it must be removed by caustic 
 soda instead of lime, or, better, by a solution of sulphate of zinc. 
 
 After the iron has been freed from sulphuric acid in the man- 
 ner just described, it is put in a bath of lime-water or milk of 
 lime, boiling-hot (it is very important that it should be hot), 
 and left there long enough to reach the temperature of the liquid. 
 It is then removed to an oven and dried, after which the lime is 
 brushed off. If desired the lime may be removed by washing 
 before putting in the oven ; in this case it will be found that the 
 surface, which is perfectly clean and bright, rusts very easily 
 and quickly; whereas if the lime is removed by drying and brush- 
 ing, the surface is much less likely to rust, although even then 
 it rusts easily and should be painted immediately. 
 
 For most of the foregoing information relating to pickling 
 I am indebted to Mr. E. G. Spilsbury, who has had extensive 
 experience in this work both in Europe and the United States, 
 and has applied the process to structural steel (bridge) work, as 
 well as to wire and wire rods. Mr. Spilsbury estimates the total 
 expense of pickling and cleaning steel for bridgework in a modern 
 and well-arranged plant at about 25 cents per ton. As to the 
 time when the cleaning is to be done, it may be said that the 
 
 34 
 
metal may be cleaned before it reaches the shop at all. This 
 is in one respect a simpler method, as requiring less handling. 
 It is obviously impracticable to do the work of cleaning either 
 by the sand-blast or acid in the shop itself, and the method speci- 
 fied involves the removal of the pieces from the shop and bring- 
 ing them back. This extra labor is variously estimated by shop 
 owners and managers who have been consulted at from 50 cents 
 to $2 per ton; doubtless some have much better facilities than 
 others; and there will also be a difference of cost in handling 
 light or heavy work. To offset this extra charge, which would 
 be, if we average the estimates of the persons consulted, about 
 $i per ton, we save the application of a shop coat of oil. The 
 metal will remain in the shop not less than a week, as an aver- 
 age; and it is necessary that it should be protected during that 
 time from rust. The shop men all agree that it will not do to 
 paint it at that time, and that an oil coat is the only thing which 
 is practicable. It is quite well known that raw linseed-oil is 
 much more durable than boiled oil, and if we apply a coat of 
 boiled oil, which will dry in about twenty-four hours, we are 
 in the irrational position of having taken great care to secure a 
 perfect surface and then applied a foundation coat of inferior 
 material. 
 
 So we are brought to the necessity of using raw oil; and this 
 will require about a week to dry, during which time the metal 
 certainly ought to be kept under cover. The expense of this 
 
 Erie R.R. bridge at Salisbury Mills, N. Y. Painted with Durable Metal 
 
 Coating. 
 
 coat of oil, applied in this manner, is variously estimated at from 
 50 cents to $i per ton; and it is pretty generally agreed that 
 the result of this method of doing the work will not be as good 
 
 35 
 
as the specified method. For it must be remembered that dry 
 linseed-oil is not a hard substance like a varnish film, but is a 
 gummy substance, which during the shop handling will get dirty 
 with solid matter and with machine-oil, which cannot be removed 
 and must subsequently be painted over; while, if we wait until 
 the shop work is done, all but the assembling, we get the sur- 
 faces perfectly clean, and can then apply our shop coat of paint, 
 which will be dry so it may be handled in forty-eight hours; and 
 then when we rivet it up we can apply a second coat of fresh 
 paint to the surfaces to be riveted together, which are by far 
 the most important to preserve, as it is in these and other crevices 
 that most of the dangerous rusting is likely to occur, especially 
 if the work is properly cared for in the way of maintenance. 
 More will hereafter be said in regard to the matter of shop coat 
 of oil. 
 
 II. Immediately after being so cleaned, and before any dis- 
 coloration due to rusting has begun, the material shall be care- 
 fully inspected, and upon the approval of the inspector, shall receive 
 one full coat of Durable Metal Coating made by Edward Smith & Co., 
 of 45 Broadway, New York City ; except planed and turned surfaces, 
 which shall be coated with Vacuum Flushing Oil and shall be 
 kept so coated until they are erected in place. 
 
 In order to carry this section into effect it will be necessary 
 to paint the work at once, within one or two hours of the time 
 when it is cleaned. If the metal has been pickled it will there- 
 fore be hot from the oven; it is then in the best condition, for 
 paint spreads better and penetrates crevices when the metal 
 is hot, and there is practically no air film on the metal to be dis- 
 placed. It is the opinion of many engineers that the best results 
 would be obtained by immersing the pieces of iron or steel in a 
 tank of hot paint, at a temperature of about 175 F., and leaving 
 it there fifteen minutes or more. By doing so all the surface 
 would be reached, and it is probable that such a course would 
 show great economy of labor, which is the most important item 
 of cost in painting. It must be remembered, whatever method 
 is adopted, that the most important matter about this whole 
 work is to get the best possible foundation coat attached as securely 
 as possible to the metal; the subsequent cost of maintenance 
 will be greatly reduced if this is effectively done. 
 
 This is the crucial point of the whole work of painting. 
 
 36 
 
III. The metal shall not be exposed to the weather nor taken 
 to the shop until in the opinion of the inspector the paint is suffi- 
 ciently dry. At no time after the preliminary cleaning shall the 
 pieces of iron or steel be laid on the ground, but shall be laid on 
 skids or trestles, and in all handling and loading and unloading 
 of the same care shall be taken to avoid scraping off the preserva- 
 tive coating; and in transportation care shall be taken to avoid 
 nesting the pieces except with packing material between them. 
 
 It is not expected that the work will be held in the paint shop 
 until the paint is thoroughly dry, but only until it has got its 
 initial set throughout; but of this the inspector is to be the judge. 
 It is probable that it will take two days, sometimes three. 
 
 IV. After being taken back to the shop all surfaces which are 
 to be there riveted or bolted together are to be cleaned, if necessary, 
 and painted on each such surface with a second coat of said Dur- 
 able Metal Coating; only such parts of surfaces as are to be thus 
 brought in contact are to be so treated. After riveting, all small 
 cavities which will hereafter be inaccessible shall be rilled with a 
 mixture of one part Portland cement and three parts clean sand, 
 with a proper amount of water. 
 
 V. After erection the work shall be carefully inspected, and if 
 there are any rusty spots these shall be thoroughly cleaned and all 
 such places and also places where the paint has been rubbed off shall 
 receive a coat of Edward Smith & Co.'s Durable Metal Coating; and 
 all exposed edges and angles shall receive an extra striping coat of 
 the same, covering the edge and the adjacent surface one or two 
 inches from the edge on each side ; all rivet- and bolt-heads and nuts 
 shall also receive an extra coat ; after this has become dry the whole 
 surface, having previously been thoroughly cleaned from dirt, shall 
 receive another full coat of^said Durable Metal Coating. 
 
 When paint begins to dry there is at first a sort of skin formed 
 on the surface, which contracts, and on rounded surfaces like 
 rivet-heads, and on angles and edges, seems to press away the 
 liquid paint beneath, so that on such surfaces there is less than 
 the normal amount. The same tendency to contract also exists 
 on flat surfaces, but in this case it is a balanced tension and pro- 
 duces no effect. There is besides the action of the painter's 
 brush, which presses harder on such places and draws off the 
 paint; but that this is not the main cause is shown by the fact 
 that pipe sections and other things which have been coated by 
 dipping exhibit the same appearance. In making paint tests, 
 
 37 
 
it is necessary to leave out of account a strip about an inch wide 
 along the edges of the plate, unless that portion has received 
 an extra coat; and the fact is well known to inspectors that such 
 
 Drawbridge at Ashtabula Harbor, L. S. & M. S. R.R. E. A. Handy, Ch. Eng. 
 Painted with Durable Metal Coating. 
 
 surfaces are always thinly coated. The extra striping coat is 
 therefore necessary if we are to have two full coats or their equiva- 
 lent over the whole surface ; and it is the more important because 
 these portions are more exposed than the flat surfaces. 
 
 VI. In no case shall a second or third coat of paint be applied 
 until the previous one is entirely dry. 
 
 VII. During erection any small cavities which will hereafter 
 be inaccessible shall be filled as provided in Section IV. 
 
 VIII. After receiving the second coat of Durable Metal Coating, 
 and after the same has become quite dry, the whole structure shall 
 receive two coats of such paint as may be designated and approved 
 by the engineer, the pigment in such paint to consist of two parts 
 by weight of white lead and one part of white zinc, tinted to a 
 designated shade; and the vehicle to be a varnish composed of 
 only hard varnish resins and pure linseed-oil and turpentine, the 
 oil being in the proportion of not less than two parts by weight 
 of oil to one part by weight of the aforesaid resin weighed before 
 melting; and the product to be free from common rosin, benzine, 
 and mineral-oil products; and samples of the paint which the 
 contractor proposes to use for this work shall be obtained not less 
 than one hundred days in advance and submitted to the engineer 
 for examination and approval, and no such paint shall be used 
 unless approved by him, and his rejection of any paint shall be 
 final; and when such sample has been approved, the contractor 
 
must use the paint made by the same manufacturer who made 
 the sample, and the paint must be, when used, exactly the same 
 in every respect as the tested sample and must not be thinned down 
 or adulterated or changed in any way whatever; and it shall be 
 used subject to all the conditions mentioned in the following section. 
 
 Section VIII is intended exclusively for structures which 
 must be painted a light color. The only pigments which can 
 be used as a base for light-colored paints are white lead and white 
 zinc; neither of these is regarded as very durable, but their use 
 is absolutely unavoidable. A mixture of the two is regarded 
 as better in many ways than either alone. They may be mixed 
 with good results in any proportion from two parts of lead to 
 one of zinc, to one part of lead to two of zinc; the former pro- 
 portion is a very common one. They are liable to be adulterated 
 with sulphate of baryta, carbonate of lime (whiting), china-clay 
 and ground feldspar. None of these has any value, and paint 
 containing any of them should be rejected. White-lead paint 
 has a tendency to show a granular surface which will brush off; 
 it " chalks," to use the painter's words; zinc tends to come off 
 in flakes. On this account it is better to use a varnish as a 
 vehicle or binder rather than oil, as the former has more hardness 
 
 Forty-inch riveted steel pipe, Cambridge, Mass., 1895. Coated with Sabin 
 
 Pipe Coating. 
 
 and tenacity; moreover, a varnish paint, if made of good mate- 
 rials, is more impenetrable to air and water, and therefore more 
 desirable. But it is important that the varnish should be suited 
 
 39 
 
to the purposes. As a rule, the larger the proportion of oil to 
 resin the more elastic and durable is the varnish; but a point is 
 reached when the resin appears to be saturated with oil, bevond 
 
 Sixty-inch riveted steel pipe, Allegheny, Pa., 1895. 
 
 Pipe Coating. 
 
 Coated with Sabin 
 
 which any more oil has a deleterious effect. The exact pro- 
 portions vary with the kind of resin, and also depend on some 
 other conditions; but in general the most durable varnishes con- 
 tain from two to two and a half times as much oil as resin, the 
 latter being weighed before melting, during which process it 
 loses about one-fourth its weight. Such a paint will not be 
 required to be perfectly white, being usually tinted to some desig- 
 nated color, and can be sold in large quantities at a price prob- 
 ably not more than 30 or 40 per cent more than a straight lead- 
 and-oil paint will cost; it will be found much more economical. 
 As to testing such a paint, it is absolutely necessary to have a 
 reasonable time, and 100 days is little enough. It would be better 
 to have more. Of course, if it is admissible to leave the struc- 
 ture black this section will be omitted; but in that case, for really 
 high-class work such as these specifications contemplate, a third 
 coat of the protective coating will be advisable. It is expected 
 that the coating of white or light-colored paint will require to be 
 renewed from time to time, while the durable metal coating is 
 depended on for permanent protection. 
 
 40 
 
IX. All paint and Durable Metal Coating used for this work 
 shall be purchased directly from the manufacturer or his authorized 
 agent, and each shipment shall be accompanied by a signed cer- 
 tificate from the manufacturer or such agent, stating that he has, 
 at that time, shipped a specified amount of the specified paint or 
 Durable Metal Coating; and all paint and Durable Metal Coating 
 shall be brought on to the premises where they are to be used in 
 the manufacturers' sealed packages, which shall be opened in the 
 presence of the inspector, who may then, and at any subsequent 
 time, take samples for examination or analysis; and in case any 
 analysis made by direction of the chief engineer shows impurity, 
 adulteration, or substitution in these specified materials, the con- 
 tractor shall pay all the costs of such analysis, and shall more- 
 over thoroughly clean off all metal coated with such impure or 
 unauthorized material and shall repaint it to the satisfaction of the 
 inspector. And the contractor shall, upon demand, exhibit to the 
 engineer or inspector the bills from the manufacturers or their 
 agents, showing the amount of paint and Durable Metal Coating 
 purchased, and also the certificates spoken of in this section; and 
 neither the paint nor the Durable Metal Coating shall be thinned 
 
 Sixty-inch riveted steel pipe, Allegheny, Pa., 1895. Coated with Sabin 
 
 Pipe Coating. 
 
 with anything whatsoever, nor shall any turpentine or benzine be 
 allowed upon the premises for any purpose, except by permission 
 of the inspector and in such quantity as he may allow. 
 
X. The inspector shall be notified when any painting, either 
 with Durable Metal Coating or other material, is to be done, and 
 no such work is to be done until the inspector has approved the 
 surface to which it is to be applied; and the contractor shall fur- 
 nish all facilities for inspection and for necessary marking by 
 the inspector, and all materials, such as paint, brushes, etc., for 
 such marking. No such inspection or marking shall be done 
 except by the engineer or his authorized inspector. 
 
 XI. In no case shall any paint be applied out-of-doors in freezing, 
 rainy, or misty weather, and all surfaces to which paint is applied 
 must be at the time dry and clean; and all work must te done in 
 a thorough, neat, and workmanlike manner. If it is necessary, in 
 cool weather, to thin the paint, this may be done only by heating 
 it; and this may be required by the inspector. 
 
 The practice of heating paint before it is applied has been 
 followed for a considerable time by some of the best engineers. 
 It is sometimes done by keeping several tin pails of paint in a 
 hot-water tank, kept hot by exhaust steam; or where there is 
 no steam plant it is heated directly over a small oil-stove or other 
 source of heat. If heated from 125 to 150, which is perfectly 
 safe if the paint contains no benzine in fact it may safely be 
 heated as hot as boiling water will make it it flows much more 
 freely, and penetrates the crevices of the work much better. It is 
 indeed likely to make a thin coat, which is objectionable; but 
 the advantages outweigh the disadvantages. 
 
 XII. The foregoing specifications shall be accepted and carried 
 out faithfully in every particular and shall not be construed accord- 
 ing to any prevalent practice not in full accord therewith. 
 
 4? 
 
1 1 ift 
 
 Fayerwether Hall, Yale University. Edward Srrith & Co.'s coatings used 
 throughout on this building. 
 
 44 
 
III. 
 
 SPECIFICATIONS FOR ORDINARY WORK. 
 
 I. Shortly before riveting, all such parts of surfaces as are 
 to be brought permanently into contact shall be thoroughly cleaned 
 from dirt and rust, and from all scale which does not perfectly 
 adhere to the metal, by the use of scrapers, chisels, and wire brushes ; 
 the latter alone shall not be considered sufficient. Each such 
 surface shall then receive one full coat of Durable Metal Coating, 
 made by Edward Smith & Co., of 45 Broadway, New York. 
 
 The wire brush is an efficient means of getting rid of loose 
 scale and dirt; but it is practically worthless for removing thick 
 rust or anything which adheres closely. Much of such material 
 may be removed by steel scrapers; but deeply corroded spots 
 should be thoroughly cleaned out with a chisel, and then well 
 brushed. These crevices are hereafter to be inaccessible; and 
 they are subject to the most dangerous corrosion, because rust- 
 ing at such places impairs not only the strength, but also the 
 stiffness, of the structure a matter of much importance. These 
 joints therefore deserve more care than any other part. 
 
 II. Shop-marks shall be compact and shall not cover more 
 surface than the inspector directs, the intent being to have the 
 surface occupied by such shop-marks as small as possible. 
 
 III. After assembling, the whole of the metal surfaces shall be 
 thoroughly cleaned in the manner described in the first section, and 
 shall then receive one full coat of said Durable Metal Coating, except 
 planed and turned surfaces and shop-marks; and all planed and 
 turned surfaces shall be coated with Vacuum Flushing Oil and 
 shall be kept so coated until they are erected in place; and all 
 small cavities which will hereafter be inaccessible shall be filled 
 with a mixture of one part of Portland cement and three parts 
 of clean sand, with a proper proportion of water. 
 
 Many engineers are in the habit of specifying a coat of either 
 raw or boiled linseed-oil at the shop, instead of a shop coat of 
 
 45 
 
paint. The advantages claimed for a shop coat of oil are, first, 
 it facilitates inspection, as the oil is transparent; second, it makes 
 an extra coat; third, the paint is likely to be more carefully 
 applied if used at the place of erection. Doubtless these con- 
 siderations are entitled to some weight; but many of the best 
 engineers practice shop-painting, which indicates that inspection 
 can be carried on efficiently after the first coat of paint has been 
 applied; and at many shops very excellent painting is done, 
 while the contractor who does the erecting is, so far as my obser- 
 vation goes, extremely indifferent to the paint question. When 
 it comes to the matter of architectural ironwork, we might as 
 well give up the idea of having two coats of paint applied during 
 
 Thirty-eight-inch riveted steel pipe, Rochester, N. Y., 1893. 
 
 Sabin Pipe Coating. 
 
 Coated with 
 
 erection; it takes too much time, and delay is a word not in the 
 contractor's vocabulary when paint is concerned. But another 
 and more important matter is this: If we apply a coat of raw 
 linseed-oil we must allow about a week for it to dry, and it cer- 
 tainly ought to be kept all this time under cover, for fresh oil 
 will wash off to a considerable extent if exposed to rain. If, on 
 the other hand, we use boiled oil, we will have it dry in twenty- 
 four hours; but boiled oil is much less durable than raw, and 
 as this is the foundation coat it is of the highest importance that 
 it should be as durable as possible; and it is much more difficult 
 to detect impurities in boiled than in raw oil. Further, there 
 has grown up among the steel men a belief that the shop coat of 
 
oil is only for protection during shipment, and with this has come 
 the practice of applying a cheap and worse than worthless sub- 
 stitute for linseed-oil, composed chiefly of mineral oil, some- 
 times further adulterated with rosin. This oil will penetrate 
 the paint which is put over it, and as the film of oil is about half 
 as thick as the coat of paint it is equivalent to thinning the latter 
 with half its volume of very deleterious and injurious matter. It 
 is then a question not how much good, but how little harm does 
 the shop coat of oil do. I have no doubt that often a good paint 
 has been a failure because of the shop coat of oil. And if the 
 shop manager cannot be trusted to apply a shop coat of paint 
 I would certainly be afraid of the oil he would use. 
 
 It is worthy of note that there is good authority for the follow- 
 ing procedure in bridge work : A shop coat of the best raw linseed- 
 oil is applied, and the work is then erected and allowed to stand 
 one to six months, until the oil is supposed to have thoroughly 
 penetrated and with the aid of the weather loosened the scale. 
 
 The whole surface is then very thoroughly cleaned (as in 
 Section I) and given two coats of paint. It is claimed that by 
 this means the mill scale is well removed, while rusting has been 
 avoided. There are reasons for thinking this good practice, 
 but it involves watchful care and thorough work. 
 
 If it is determined to use oil, the following specification is 
 recommended : 
 
 The oil used for this purpose shall be raw linseed-oil of the 
 best quality; shall be permanently clear and settled by aging 
 or storage for not less than thirty days ; free from rosin as well as 
 mineral-, fish-, and cottonseed-oils, or any other impurities; shall 
 contain no benzine, turpentine, or other solvent, nor any lead 
 or manganese compounds; it shall be purchased directly from 
 some reputable manufacturer of linseed-oil, and each shipment 
 shall be accompanied by a signed statement from the manufac- 
 turer saying that he has, at that time, shipped a specified amount 
 of pure raw linseed-oil of the best quality ; and it shall be brought 
 on the premises where it is to be used in the manfacturer's sealed 
 packages, which shall be opened in the presence of the inspector, 
 who may then and at any subsequent time take samples for exam- 
 ination or analysis ; and in case any analysis made by direction 
 of the chief engineer shows impurity, adulteration, or substitu- 
 tion in this material, the contractor shall pay all the costs of 
 such analysis, and shall moreover thoroughly clean off all metal 
 coated with such impure or unauthorized oil and recoat it to the 
 satisfaction of the inspector. And the contractor shall upon 
 
 47 
 
demand exhibit to the engineer or the authorized inspector the 
 
 bills and statements from the manufacturer showing the amounts 
 
 of oil purchased. 
 
 In case boiled 'oil is to be used the word " boiled " would be 
 
 , substituted for " raw " in the foregoing, and 
 
 omit "nor any lead or manganese com- 
 pounds." 
 
 But the writer of this is of the opinion that 
 a shop coat of oil is in most cases a useless 
 expense (which might far better be applied to 
 a striping coat), and in a large proportion of 
 cases a positive and great injury. Vacuum 
 flushing oil is a very heavy mineral oil, about 
 as heavy as an oil for wagon axles, and has 
 been successfully used for a long time. It 
 appears to be much better than the usual 
 mixture of white lead and tallow, which is not 
 very satisfactory. It is not expensive. 
 
 " All small cavities." It is not proposed, 
 for example, to fill hollow columns in this way. 
 Large cavities which are liable to be wet 
 should be drained; but this is not a matter of 
 painting. 
 
 Unlined steel stack, 
 310 feet high, of 
 the Nichols Che- 
 mical Co., Laurel 
 Hill, N. Y. City. 
 This has stood 
 six years, paint- 
 ed on erection 
 with Durable 
 Metal Coating 
 inside and out- 
 side, without re- 
 painting. 
 
 IV. The metal shall not be exposed to the 
 weather nor loaded for shipment until in the 
 opinion of the inspector the paint is sufficiently 
 dry. At no time after the application of the 
 first coat of paint shall the pieces of iron or 
 steel be laid on the ground, but shall be laid on 
 skids or trestles ; and in all handling and loading 
 or unloading of the same care shall be taken 
 to avoid scraping off the preservative coating; 
 
 and in transportation care shall be taken to avoid nesting the 
 
 pieces except with packing material between them. 
 
 This section calls for shop painting to be done under cover. 
 This is probably the most radical reform called for in these speci- 
 fications. Few shops have a paint-shop where such work may 
 be done; but while no serious objection can be raised to painting 
 out-of-doors in dry, warm weather, it must be remembered that 
 this kind of work is going on all the year round, and that during 
 a large part of the time the weather is such that it is impossible 
 
 48 
 
to do good work outside. It seems to the writer that if this 
 single reform can be enforced, as it is in all other kinds of paint - 
 i n g> a great advance will be made in the matter of preserving 
 steel from corrosion. 
 
 Incidentally it has the effect of making the painting a dis- 
 tinct step in the shop work, and recognizing paint as proper 
 engineering material. The rest of the section is of the same 
 tenor, and will be found difficult to enforce, not because it involves 
 any hardship or much expense to the contractor, but because 
 steel is handled in a slovenly and careless manner, being usually 
 thrown off wagons or cars into the mud; and inspectors have 
 really very great difficulty to get the mud cleaned off, or even 
 to have a pretence made of cleaning it off, before the second coat 
 of paint is applied. The reform needed is a change of the spirit 
 in which the work is done, and must lead to a recognition by all, 
 from the owner to the humblest laborer, of the fact that steel 
 is a perishable material the preservation of which is a matter 
 of serious and constant importance. 
 
 V. After erection the work shall be carefully inspected, and if 
 there are any rusty spots these shall be thoroughly cleaned, and 
 all such places and also all places 
 where the paint has been rubbed 
 off shall receive a coat of Edward 
 Smith & Co.'s Durable Metal Coat- 
 ing; and all exposed edges and 
 angles shall receive an extra striping 
 coat of the same, covering the edge 
 and the adjacent surface one or two 
 inches from the edge on each side ; 
 all rivet- and bolt-heads and nuts 
 shall also receive an extra coat; 
 after this has become dry, the whole 
 surface, having previously been 
 thoroughly cleaned from dirt, shall 
 receive another full coat of said 
 Durable Metal Coating. 
 
 Steel stack, 200 feet high, of the 
 Philadelphia Smelter, Pueblo, 
 Colo. Painted with Durable 
 Metal Coating. 
 
 VI. In no case shall a second or 
 third coat of paint be applied until 
 the previous one is entirely dry. 
 
 It is proper at this point to call attention to the fact that 
 some engineers specify that successive coats of paint shall differ 
 
 49 
 
in color; some advise this on the ground that it makes it easier 
 for the workmen to see if they are missing any places, and any- 
 thing which makes painting easier is likely to be advantageous; 
 others say it facilitates inspection and enables the engineer to 
 find out for himself if the required number of coats has been 
 applied. On the other hand, some say there is no difficulty in 
 distinguishing fresh paint from that a day or two old that, in 
 fact, any inspection worthy of the name will determine whether 
 the work is properly done without regard to this, while com- 
 petent inspection of bridges cannot be done from the platform of 
 a railroad car. Some of the most experienced and careful 
 engineers specify paints of different colors for different coats, 
 but others of equal standing do not approve of it ; as a result of 
 extended personal inquiry I am satisfied that a great majority 
 belong to the latter class, and as I myself hold this opinion I have 
 not specified such difference in colors, but only call attention 
 to the matter, so that those who wish may insert an article to 
 that effect. 
 
 VII. During erection any small cavities which will hereafter 
 be inaccessible shall be filled as provided in Section III. 
 
 VIII. All Durable Metal Coating used for this work shall be 
 purchased directly from the manufacturer or his authorized agent, 
 and each shipment shall be accompanied by a signed certificate 
 from the manufacturer or such agent, stating that he has, at that 
 time, shipped a specified amount of the Durable Metal Coating; 
 and all Durable Metal Coating shall be brought on to the premises 
 where it is to be used in the manufacturer's sealed packages, which 
 shall be opened in the presence of the inspector, who may then, 
 and at any subsequent time, take samples for examination or 
 analysis; and in case any analysis made by direction of the chief 
 engineer shows impurity, adulteration, or substitution in these 
 specified materials, the contractor shall pay all the costs cf such 
 analysis, and shall moreover thoroughly clean off all metal coated 
 with such impure or unauthorized material and shall repaint it 
 to the satisfaction of the inspector. And the contractor shall, upon 
 demand, exhibit to the engineer or inspector the bills from the 
 manufacturers or their agents, showing the amount cf Durable 
 Metal Coating purchased, and also the certificates spoken of in this 
 section; and the Durable Metal Coating shall not be thinned with 
 anything whatsoever, nor shall any turpentine or benzine be allowed 
 upon the premises for any purpose, except by permission of the 
 inspector and in such quantity as he may allow. 
 
IX. The inspector shall be notified when any painting is to be 
 done, and no such work is to be done until the inspector has approved 
 the surface to which it is to be applied; and the contractor shall 
 furnish all facilities for inspection and for necessary marking by 
 the inspector, and all materials, such as paint, brushes, etc., for 
 such marking. No such inspection or marking shall be done 
 except by the engineer or his authorized inspector. 
 
 X. In no case shall any paint be applied out-of-doors in freezing, 
 rainy, or misty weather, and all surfaces to which paint is to be 
 applied must be at the time dry and clean; and all work must be 
 done in a thorough, neat, and workmanlike manner. If it is 
 necessary in cool weather to thin the paint, this may be done only 
 by heating it, and this may be required by the inspector. 
 
 XI. The foregoing specifications shall be accepted and carried 
 out faithfully in every particular and shall not be construed accord- 
 ing to any prevalent practice not in full accord therewith. 
 
 5' 
 
_d o 
 7D w 
 
IV. 
 SOME ANTI-CORROSIVE COATINGS. 
 
 Edward Smith & Company manufacture all lines of such 
 varnishes as are used in structural work, as well as varnishes 
 for carriages, railway coaches, yachts, and the like; but besides 
 these they have for many years made a special business of making 
 protective coatings to prevent the corrosion of iron and steel 
 in bridges, train-sheds, water-tanks, and mains, and in fact all 
 work where metal is exposed to risk of corrosion. There is no 
 one material which is best for all places ; and they maintain a 
 chemical and testing laboratory, well equipped, and their officers 
 are in a position to offer intelligent and expert advice on almost 
 all questions of this sort, and to make special preparations for 
 unusual conditions. This is now so well known that they are 
 consulted as experts by many engineers of the highest repute. 
 
 Durable Metal Coating. 
 
 This is the best known and most popular of these anti-rust 
 preparations made by this company, and the following is a brief 
 account of its qualities and uses. 
 
 It is an oleo-resinous varnish, of the highest durability and 
 elasticity; it is black in color. When dry it is not affected by 
 nor does it affect any paint put on over it ; it is therefore practic- 
 able to use paint of any color as a finishing coat if desired. Two 
 coats of white zinc or white lead cover well, and if the desired 
 color is of a neutral or dark shade one coat is often sufficient. 
 If it is desired to throw it off color only a little, the painter may 
 add to it pigments either dry or ground to a paste in oil; for 
 example, a bronze-green suitable for certain ornamental ironwork 
 in^parks, etc., may be had by adding a little chrome green, or 
 olive by medium chrome yellow. For engineers who desire to 
 use different colors in successive coats, we will make it in dark 
 green, chocolate-brown, or olive, if the order is of considerable 
 
 53 
 
amount. Used without the addition of any color it gives a very 
 handsome and permanent black finish to iron and steel, which is 
 generally liked, black being commonly regarded as a natural 
 
 Steel beams (painted with Durable Metal Coating) of the Stadium, Harvard 
 University. Prof. L. J. Johnson, engineer. 
 
 and proper color for iron. Most of it is used in this 
 way. 
 
 One gallon will cover from 300 to 400 square feet, according 
 to the character of the surface. If the latter is very rough, owing 
 to corrosion, it may not cover more than 250 square feet the 
 first coat, because in that case it acts as a filler, but it will cover 
 more than 300 square feet for the second coat. There is not a 
 great difference in the amount of surface covered by different 
 paints, if the surfaces are alike in material and smoothness, because 
 all paints and varnishes are made up to a certain conventional 
 standard of viscosity, suitable for working with the brush; and 
 any paint or varnish made for general use must be modified if 
 used in an unusual way, if applied by dipping, for example; and 
 a skilful painter may make any structural paint cover 500 or 
 600 square feet by spending enough labor on it brushing it out 
 thin; but this is practically not done because the labor costs 
 more than the paint, and theoretically it never should be done, 
 because such a coating is too thin to afford sufficient protection. 
 For decorative effect thin coats are best; for protection they 
 should be as thick as can be applied uniformly and lie smoothly. 
 
 54 
 
Paints may be made to cover more surface if they are made very 
 thin with benzine or turpentine, but are then of little value. 
 
 Durable Metal Coating is supplied ready to use, and must 
 not on any account be thinned with anything. It is a perfect 
 liquid, containing no pigment, does not require stirring, and 
 can be used to the last drop in the barrel. It contains no coal- 
 tar, no naphtha or other coal-tar product; no benzine or any 
 petroleum derivative; no rosin or rosin-oil; no materials but 
 the purest and best for the purposes for which it is made. It is 
 usually sold in barrels, but will be put up in packages of any 
 size, smaller packages being charged extra at the usual rates. 
 It may be shipped to India and other hot countries in ico-gallon 
 drums, or in sealed cans, if barrels are not thought suitable on 
 account of the climate. 
 
 It is not necessary to have skilled labor for its application, 
 unless it is to be used as an ornamental coating, as it sometimes 
 is. If it is necessary (as often happens) to use it out-of-doors 
 in cold weather, it mav be heated to make it more fluid. Thou- 
 
 Pierce Hall (Engineering building), Harvard University. Steel framework 
 painted with Durable Metal Coating. Prof. L. J. Johnson, engineer. 
 
 sands of gallons have thus been treated by the Northern Pacific 
 Railroad Company and others; as it contains no benzine it is 
 perfectly safe to heat it to as high a temperature as it is reason- 
 
 55 
 
able to subject the bristle brushes which are used in applying 
 it say to 150 F. 
 
 It is commonly estimated to require about thirty-six hours 
 to dry so that it may be handled, and from three or four days 
 to two weeks for it to dry through. It does, however, continue 
 to harden for at least two weeks, and in cold or wet weather it 
 may require more than thirty-six hours before it may be handled, 
 while in hot, dry weather it has been known to dry to this extent 
 in from eight to twelve hours. 
 
 Pierce Building (for engineering), Mass. Institute of Technology, 
 framework painted with Durable Metal Coating. 
 
 Steel 
 
 A second coat should not be applied in less than three days 
 and a week or ten days is much better. In fact, in bridgework 
 it is good practice to allow two to four weeks between coats, with 
 any kind of paint. But if it can be dried by heat, as is the case 
 with much .ornamental ironwork and with small objects gen- 
 erally, it will dry in two or three hours at 200 to 250. It dries 
 by oxidation, and heat accelerates this process. If it is impossible 
 to allow a sufficient time between coats, we advise the use of our 
 Ebonite varnish for a first coat, which* dries much more quickly. 
 
 We do not give any guarantee' of performance. No reputable 
 manufacturer can afford to do this, not knowing the character 
 of the surtace to which the paint is to be applied, the intelligence 
 
 56 
 
of t he-man who applies it, the treatment it is to receive, or any 
 of the many things which affect the durability of a paint or var- 
 nish. If the reader will take the trouble to look up with the 
 aid of any of the commercial agencies the standing ot the paint 
 manufacturers, he will find that no one having a considerable 
 rating offers a guarantee which amounts to anything; some- 
 times, to effect a sale, a guarantee is offered, but careful exam- 
 ination will always show that it is so protected by restrictions 
 that no real guarantee is given. In general, no paint guarantee 
 is^good for anything except a contract to keep a structure painted 
 a specified time for a certain sum of money. The curious inves- 
 tigator will also notice that some of the concerns which give 
 recommendations for certain paints are not mentioned in the 
 books of the commercial agencies; and that sometimes pictures 
 of structures are sent out with advertising matter which were 
 not painted with the advertiser's material; it is not expressly 
 claimed that they were so painted, but the obvious intention 
 is to make the public believe it. There are plenty of reputable 
 paint manufacturers, and it is not necessary to do business with 
 tricksters. 
 
 Edward Smith & Co. wish to call attention, especially on 
 the part of architects and engineers, to the fact that it is desir^ 
 able to specify " Durable Metal Coating made by Edward 
 Smith & Co. of 45 Broadway, N. Y.," because at least one con- 
 cern has sent its agents around to our customers offering to sell 
 them a paint under the same name and guaranteed " equal to 
 Smith's," at prices varying from 90 cents or $1.00 per gallon, 
 in the case of architects, to as low as 12^ cents a gallon to the 
 large steel companies. This paint was not worth 12^ cents per 
 gallon, and will not be satisfactory to people who want Edward 
 Smith & Co.'s goods; and it has been used when the latter were 
 specified. Therefore take warning. 
 
 While we strongly recommend thorough cleaning of all sur- 
 faces which are to be painted, we do not wish to be understood 
 as meaning that our paint cannot be applied to any but a clean 
 metallic surface. It has been and is constantly being used as a 
 protective coating over red lead (on the ground that the good 
 qualities of the latter depend on its protection from the atmos- 
 sphere); and there is no serious objection to using it over old 
 paint which is in good condition and adheres well. 
 
 Durable Metal Coating has a considerable use in railroad 
 work. The bridge at Eagle Bridge, N. Y., consisting of two 
 plate-girder spans aggregating about 200 feet in length, was 
 
 57 
 
painted with two coats of this material when erected, in 1893; 
 it gets all the refrigerator drip from the east-bound trains of 
 the Fitchburg R.R. ; it has not [been repainted, even in part, 
 
 Eagle Bridge, N. Y. Painted with Durable Metal Coating. 
 
 and is now (1905) in excellent condition. Now, it is impossible 
 to make a paint which will stand refrigerator drip ten or twelve 
 years unless it is carefully and thoroughly applied to new or 
 clean metal, under favorable conditions; and any good paint 
 thus applied will give better than average results ; but it is beyond 
 question that a paint which has stood as long as this, and is appar- 
 ently good for many years yet, must be of exceptional quality. 
 _JTo take another case: In 1896 the large train-shed of the 
 Illinois Central R.R. in Chicago (which is also used by the Michi- 
 gan Central) was -painted inside with Durable Metal Coating, 
 two coats. This was a repainting job; the paint originally used 
 Kkd perished, and was thoroughly cleaned off by scraping and 
 wire-brushing before the application of our material. This stood 
 all right without repairs for six years ; at the end of which time 
 the exterior of the shed was being painted with Durable Metal 
 Coating, and a third coat was then put on the interior; no scrap- 
 ing was necessary, only ordinary cleaning with brushes; the 
 1*896 paint was all right, but the additional coat was applied 
 over the others. It is now in excellent condition. This is both a 
 terminal- and a through- train shed, 670 by 140 feet, and the test 
 is a severe one. Its fine condition was one reason why the Chicago 
 
 58 
 
& Northwestern R.R. painted their large train-shed at their 
 main station (Wells Street) in Chicago, in 1903, with Durable 
 Metal Coating. This is the only C. & N. W. structure of con- 
 siderable importance, so far as known to the writer, which is 
 painted with anything but the C. & N. W. standard paint made 
 in the shops of the railroad company; it was adopted without 
 solicitation, and after careful investigation by the officials in 
 charge of the work. 
 
 The Illinois Central has used thousands of gallons of Durable 
 Metal Coating for bridge work and are still using it, especially 
 for first-coat work, at the bridge shops. A large amount has 
 been used on the Northern Pacific and other roads in the North- 
 west; this was probably the reason that caused the Canadian 
 Pacific R.R. to use it on their train-shed at Montreal and else- 
 where; they were never asked to use it. 
 
 Another thing for which Durable Metal Coating is used is for 
 painting wire cables of electric wires, such as are used for block- 
 signal lines and the like. The New York Central R.R. has used 
 
 Highway Bridge at Albany, N. Y. Painted with Durable Metal Coating. 
 Stowell and Cunningham, designers. 
 
 some thousands of gallons for this purpose; and the Western 
 Union Telegraph Co. has used considerable quantities of it for 
 many years. It is also used by the Postal Telegraph Co. on 
 
 59 
 
cables and other work. It has proved satisfactory for painting 
 all sorts of exposed work for telegraphs and electric light and 
 power, such as metallic poles, cross-arms, and the like. For all 
 
 Train-shed of the Wells Street Depot of the C. & N. W. R.R., Chicago. 
 Painted, 1903, with Durable Metal Coating. E. C. Carter, Ch. Eng. 
 W. H. Finley, Prin. Asst. Eng. 
 
 such uses it is better than an oil paint, because, for one thing, 
 it contains no pigment; most sorts of pigment are conductors 
 of electricity, and the better insulator a coating is the less likely 
 is rusting to be set up, particularly on metal which is more or 
 less charged with electricity. The metal work of steel-frame 
 buildings is essentially similar to bridge work, and many of the 
 finest structures of this sort in the country have been protected 
 with this coating. If it is necessary to use a more rapid drying 
 material our coating called Ebonite is recommended. 
 
 For painting water-towers and tanks for drinking-water it 
 is unequalled. It is generally admitted that^lead paints are not 
 suited for such uses, because of the possibility that the lead may 
 be dissolved in the water; in this varnish there is nothing which 
 can be dissolved, nor that would be objectionable if it were dis- 
 
 60 
 
solved. It is particularly liked for painting tanks for water 
 on ships and yachts; it never becomes loosened by the vibra- 
 tion of the boat. In his annual address one of the presidents 
 of the New England Water Works Association mentioned it as 
 the best coating for water-towers; a great many of them have 
 been painted with.it, and we have yet to have the first complaint. 
 It has also been used for ten years on both steel and cast-iron 
 water-mains, with satisfactory results. 
 
 It is well adapted for painting metal roofs. It is well known 
 that new tin or galvanized-iron roofs are difficult to paint; this is 
 because in making the sheets of coated metal they receive a closely 
 adherent thin film of such sort that paint of any kind will not 
 adhere to it; the remedy is to thoroughly clean by scrubbing 
 with soap and water, and if greasy by rubbing with coarse cloths 
 wet in benzine. If the roof is very thoroughly cleaned any paint 
 will stick on it. Durable Metal Coating is extremely water-proof 
 and permanently elastic, and is in every way a good roof paint. 
 
 Windsor Street Station, Montreal, of the Canadian Pacific R.R. 
 painted with Durable Metal Coating. 
 
 Train-shed 
 
 As it stands high temperatures well, it is suitable for painting 
 smoke-stacks, steam-pipes, and the like. No organic material 
 will stand a heat much in excess of 400 F.; beyond this it is more 
 
 61 
 
or less rapidly burned off; but this coating bakes to a sort of 
 enamel with any reasonable temperature, and is a very per- 
 fect coating. It is not affected by acid gases and is extensively 
 
 Eye-bar chains in the anchorages of the Williamsburg Bridge. Protected 
 with Ebonite Varnish. 
 
 used about chemical works; the chief cause of deterioration of 
 paint in such places is the action of acids pn the pigment which 
 constitutes a large part of an ordinary paint; as this does not 
 contain any pigment it is not affected. 
 
 It stands well over sea- water, not being readily affected by 
 the spray; but we do not recommend it for use under the sur- 
 face of sea-water. If Durable Metal Coating is to be used on 
 decorative work, it must be brushed out thin; otherwise it will 
 wrinkle. If it is brushed out thin it dries more rapidly, which 
 is an advantage for this kind of work; but enough coats must 
 be applied to make a substantial film. If treated intelligently 
 in this way it has been a most satisfactory material; for example, 
 it was thus used on the ornamental ironwork of the Delmonico 
 building, by John Williams, well known as a leading manufac- 
 turer of bronze and iron ornamental work, in whose shops it has 
 long been used. But it must not be forgotten that a very thin 
 
 62 
 
coat of the best material will not long stand exposure to the 
 weather; enough must be used to ensure protection. 
 
 Ebonite Varnish. This is a preparation which differs from 
 Durable Metal Coating by containing less oil in proportion to 
 the resinous material; it dries more rapidly on that account, 
 and its drying is greatly increased by the fact that as it contains 
 less oil and would therefore be less fluid it is necessary to make 
 it contain more turpentine; and a large proportion of turpen- 
 tine makes a varnish dry quickly. It is subject to the objec- 
 tion that it makes a thinner film than the other; but besides the 
 advantage of quicker drying it naturally resists alkalies bette r 
 because of its large proportion of resinous material; this latter 
 being unaffected by alkali. Where the coating is to be in con- 
 tact with hydraulic cement this is an important advantage. The 
 entire metal work in the anchorages of the Williamsburg bridge, 
 connecting New York and Brooklyn, was painted with several 
 coats of this compound; and its good condition several years 
 
 Eye-bar chains in the anchorages of the Williamsburg Bridge, N. Y. City. 
 Protected with Ebonite Varnish. 
 
 later was undoubtedly one reason why the equally important 
 task of making a coating material for the cables of that great 
 engineering work was entrusted to our company. 
 
 63 
 
Another instructive instance of its use in contact with cement 
 is afforded by a paper read before the American Institute of 
 Mining Engineers in February 1904, describing the dust-flues 
 at the smelter in Leadville, Colo., built in 1898^ and still in good 
 condition. The engineer of construction said: 
 
 " In regard to the effect of sulphur dioxide and furnace-gases 
 on the cement, I have found that in certain cases this is a matter 
 which must be given very careful attention. When there is 
 sufficient heat to prevent the existence of condensed moisture 
 inside of the flue, there is apparently no action whatever on the 
 cement, but if the concrete is wet, it is rapidly rotted by these 
 gases. At points near the furnaces there is generally sufficient 
 heat not only to prevent internal condensation of the aqueous 
 vapor always present in the gases, but also to evaporate water 
 from rain or snow falling on the outside of the flue. Further 
 along a point is reached where rain-water will percolate through 
 minute cracks caused by expansion and contraction, and reachs . 
 the interior even though internal condensation does not occur 
 there in dry weather. From this point to the end of the flue 
 the roof must be coated on the outside with impervious ma- 
 terial. In very long flues a point may be reached where 
 moisture will condense on the inside of the walls in cold weather. 
 From this point to the end of the flue it is essential to protect 
 the interior with an acid-resisting paint, of which two or more 
 coats will be necessary. For the first coat a material contain- 
 ing little linseed-oil is best, as I am informed that the lime in 
 the cement attacks the oil. For this purpose I have used Ebonite 
 varnish, and for the succeeding coats, Durable Metal Coating." 
 
 Ebonite is" particularly serviceable for work on steel-frame 
 buildings, especially where haste is unavoidable. It dries so 
 that it may be handled in a few hours, and two coats may be 
 applied in a day; but the rapidity of drying is somewhat more 
 apparent than real ; what really happens is that it takes an unusu- 
 ally hard preliminary set, and the final permanent drying is about 
 as slow as that of Durable Metal Coating. Having' had^ rt in 
 use about ten years, we do not hesitate to recommend it. 
 
 Edward Smith & Company have made a great variety of spe- 
 cial protective and decorative coatings for particular uses. For 
 instance, in a case where it was desired to paint a highway bridge 
 a certain color, they made a varnish enamel paint especially for that 
 bridge. This was made with a special varnish, adapted to resist 
 the action of the weather and to protect the iron ; some varnishes 
 do not act well on iron, and no ordinary varnish will permanently 
 
resist the weather in such a place as a bridge over sea-water. 
 This was the bridge connecting the city of Boston with Charles- 
 town, the one on which the elevated railway crosses; and it is 
 subjected to very trying conditions, being exposed to smoke 
 from tugs and other vessels, to fog and spray from sea-water, 
 and great extremes of rapidly varying temperature. After five 
 or six years' use, during which time not the least repainting has 
 been necessary, the condition of the surface is such as to con- 
 stitute the best sort of a recommendation for the paint. This 
 is one of the largest and most important structures of its kind 
 in the country; and when we consider the high grade of municipal 
 work done in Boston, the fact that the coating is satisfactory 
 to the engineers for the city is enough to recommend it. 
 
 The baked enamel coating of our invention which has been 
 so successfully applied to pipework and, as described in the first 
 part of this book, to the floor of the Williamsburg bridge, and 
 ihe special coating made for the cables of that structure, are 
 other instances of our special work. It may be confidently stated 
 that no other company has done anything like as much of this 
 difficult special work, or has the scientific and technical equip- 
 ment and experience to successfully carry it out. This is a con- 
 sideration which should properly influence engineers and archi- 
 tects in the choice of material. 
 
 65 
 
 OF THE 
 
 UNIVERSITY 
 
 OF 
 
UNIVERSITY OF CALIFORNIA LIBRARY 
 
 THIS BOOK IS DUE ON THE LAST DATE 
 STAMPED BELOW 
 
 OCT 3 1916 
 161918 
 
 AUG 24 1925 
 
 DEC 1@ 1929 
 ' 2NQV-64BE 
 
 REC'D 
 
 OCT19'64- 
 
 LD 
 
 1PM 
 
YC 1826