I In One Volume : PART I. — Fundamentals. PART 2. — Finer Painting. fainting and i^ainter^^ MaferiaU: A Book of Facts for Painters and those who use OR DEAL IN Paint Materials. Treating of Oils in all their relations to Paint and Colors ; of Pigments, their qualities, uses, changes, adul- terations, AND tests ; Varnishes, their materials, compar- ative qualities, uses in decoration, and their mysteries AND CHANGES IN USE ; OF DRIERS AND THEIR EFFECT IN THE DRYING OF PAINT AND VARNISH; OF WoOD AND IrON AS PRE- served by paint, and their relations to cracking and peeling of paint and varnish; of the management of Paint Shops, Carriage Painting and Car Painting; of Decoration and the use of Color; and of the effects OF paint on health. By CHARLES L. CONDIT, UNDER THE SUPERVISION OF JACOB SCHELLER, MASTER PAINTER. PUBLISHED BY The Railroad Gazette, 73 Broadway, NEW-YORK. 1883. Copyright, 1882, by The Railroad Gazette. Atkin & Pkout, Piiutcrs, 12 Barclay Street, New York KLECTBOTYPKD BY A. IIINGLER & C NEW YORK. PREFACE. Skillful brushing and coloring do not come by read- ing books ; printed words at best furnish a clumsy method of showing the nice manipulation of a mechan- ical operation. Such teaching is better left to the shop; but in the shop one will not always acquire requisite knowledge, and will seldom or never get it formed and definitely shaped, as a printed record may shape and form it. Hence the opportunity of a text book, which shall attempt to do what the shop leaves un- done. It is thought necessary to speak of what may be called scientific aspects of the subject treated in this little volume. Its point of view is here also practical, technical statements (tables, measurements, etc.), are used to give definiteness to practical certainties. Three questions will indicate the method of investi- gation : What is the practical evidence ? What has technical investigation discovered or made probable ? Do the results of technical investigation fit the con- clusions of large experience ? Mr. Scheller is, in fact, joint author of this book ; it is also indebted to Mr. Drummond and to Mr. Day for valuable assistance in the preparation of the sections bearing also their names. Any responsibility on their part is, however, limited to these sections. To G. J. Mulder's Die Chemie der austrocknenden Oele is due the ability to give a clear exposition of the process of drying and to show wherein the quality of one paint differs from that of another. The tables of 7n Faintuig and Painters' Materials, o\\ drying," are due to Mulder, where not attributed to Chevreul, whose work has also been helpful. In view of the practical evidence, the writer has restated positively some points implied rather than clearly stated by Mulder. J. G. Gentile's Lehrbuch der Farben-fabrikation has been the source of much in- formation about pigments. Many debts are also due Notes on Building Construction, Von Bezold's Theory of Color (translation) and especially to Prof. Rood's Modern Chromatics. At the close of the volume the reader will find a brief bibliography. CONTENTS. PART I.— FUNDAMENTALS. CHAPTER I. Things in General. Page. The Object and Effect of Painting Wood and Iron i CHAPTER II. The Cracking of Varnish and Paint. Cracking, Peeling and Perishing — Decay of Picture Varnish — Expan- sion and Shrinking of Woods— Effect of Heat— Cracks Across the Grain; With the Grain; on Carriages, Cars, Interiors and Exteriors of Buildings, and on Furniture — Causes of Cracking, and one Preven- tive 6 CHAPTER III. The Purchase and Manufacture of Varnish. The Resins used, their Qualities and Prices — Mucilage and Resins — English and American Varnishes — Qualities and Tests 19 CHAPTER IV. The Use of Varnish. Effects to Obtain and Avoid — Use of Rough-stuff and Rubbing Varnish. 27 CHAPTER V. Testing the Qualities of Common Pigments. Covering Power, what it Consists of — General Test for White Paints — Kind and Quality of Color — Durability of Different Pigments — White Lead Adulterations 33 CHAPTER VI. Priming Wood. Composition and Structure of Wood — Nature of the Connection be- tween Paint and Wood — Method and Object of Priming — Peeling off from Primer 50 CHAPTER VII. Oils. A Little Dictionary " of Oil Terms — What is an Oil? — Soaps and Oil Acids 56 vtit Fainting and Painters' Materials, CHAPTER VIII. The Drying of Oils and the Drying Oils. Changes in Oils in Drying— Changes in Weight— Influence of Darkness, Light and Colored Light— Influence of Heat— What is Lost and Gained in Drying— Glycerine Ether— Freezing Temperature of the different Drying Oils— Description of the Flax Plant and Seed— How Linseed Oil is Made— Description and Qualities of the Drying Oils. . . 6i CHAPTER IX. Old Oil and Fatty Oil. Some Old Oils dry faster than New— Quick-drying Old Oil, Fatty Oils — Oxy-linseed-oil Acid— Experiments with Oleic Acid and Linseed Oil. Effect of Not-Drying Oils on Paint 8i CHAPTER X. Driers of Drying Oil. Experiments with Linseed Oil, White Lead, Zinc White, Warmed Oil, Iron Oxide, Litharge and Red Lead — Lead and Oxygen — The Oxy- gen of Driers — Boilin Oil — Changes in Oil under Heat — Raw and Boiled Linseed Oil Soaps 87 CHAPTER XI. The Causes of Decay in Paint and Varnish. Perishing — Wearing Away — Cracking, Want of Elasticity, Due to Con- version of too much of its Acids into Soaps, to too much Drier, to Losing Non-drying Oil Acid at the Wrong Time, and to Unequal Amounts of Oil in Adjoining Coats — Peeling — Sun-cracked Varnish Breaks Across the Grain, Water-cracked Varnish with the Grain — The Cracking of Paintings 96 CHAPTER XII. Manufacture and Use of White Lead. Dutch, French and German Processes — Qualities of Durable Paint — What is White Lead ? — Powdering — Adulterations — Practical Use of this Information xo8 CHAPTER XIII. Rust. Different Kinds — Tinned and Galvanized Iron — Causes of Rapid Rust- ing — Experiments — Chemistry of Rust — Kinds of Rust — Relations of Parts of an Iron Structure — Protections — Paint on Different Metals. . 124 CHAPTER XIV. Painting Iron. Cast and Wrought Iron and Steel — Priming — Red Lead — Coal Tar and Asphaltum — Practical Experience 137 Contents. ix CHAPTER XV. Paints for Iron and for General Out-Door Use. White Lead- Iron Paints — How Sunlight Injures Paint — How Pigments Protect the Oil— Red Lead — Most Economical Paint for Iron and Wood 151 CHAPTER XVI. Brushes. Qualities Required — Hog's Bristles — Adulteratiorfs — Bristle Brushes — Finer Brushes — Fitch, Skunk and Camel's (Squirrel) Hair — Use and Care 163 CHAPTER XVII. Putty, Paste and Glue. Different Materials in Putty— Drying — Rules for Use— Panel, Glycer- ine and Water-Glass Putty and Putty for Porcelain — Paste — Glue 168 CHAPTER XVIII. The Effects of Paint on Health. Lead Poisoning of Painters, Manufacturers and Others — Statistics of Death Rates — Effects of Fresh Paint — Effects of Poison — Symptoms — The Pulse in Lead Colic —Preventives 173 PART II— FINER PA IN TING. CHAPTER I. Linseed Oil and the Painter's Difficulties with it. A Summary of Facts concerning Linseed Oil and its Use 191 CHAPTER n. The Eccentricities of Oil and Varnish. Due to Atmospheric Causes — to Unclean Surfaces — to Quality of Var- nish — Mysterious Causes .. 214 CHAPTER III. The Examination of Linseed Oil and of Turpentine. ■* Tests for Adulterations of Linseed Oil — Use, Manufacture and Tests of Turpentine 219 CHAPTER IV. The Color of Linseed Oil and Changes in Varnish. Causes, Tendencies and Results of Changes in Color — Effects of Alka- lies — Green Carriages — Practical Conclusions — Bleaching of Oil — Cause and Cure of Whitening of Varnish — The Ultramarine Disease. 229 X Painting and Painters Materials. CHAPTER V, The Carriage Varnish Shop. Requirements — Location — Walls— Size — Windows — Heating — Ventila- tion — Thermometer and Barometer — A Plan — Furniture — Painting a Carriage, Color Coats and Varnishing — Good Rules 246 CHAPTER VI. System in Car Shop. Organization and Accounts— Piece Work— Painting a Passenger Car, two Methods — Repainting — Removing Paint and Varnish — Repaint- ing over Old Paint — Interior Painting 268 CHAPTER VII. Description and Test of Pigments, Composition and Qualities of all Pigments in Use (see Index to Pig- ments in pages following) — The use of Pigments, a Decorator's Ad- vice, an Artist's Advice — Preparation for Use, Color-Grinding, Mix- ing, Washing 293 CHAPTER VIII. The Chemical Relations of Pigments. Changes in Color due to Sulphureted and Ammonia gas. Oxygen, Chalk, Lime, to Mixture and to Effects of Light — Poisonous Water- Colors — Organic Compounds as Colors — Tests and Protection of Colors — Hot Colors 367 CHAPTER IX. Decoration by Color. General Principles — Harmony — Contrast — Good Combinations — Changes of Hue Due to Contact — Practical Rules and Examples for Decorating Buildings, Blinds, Doors, Walls, Floors, Cars and Car- riages, and Methods of Painting Them— Gilding — Stenciling — Draw- ing — Coloring 383 APPENDIX. Bibliography 449 Restoration of Paintings 451 Lead Poisoning — Personal Advice 452 Eye Exhaustion and Color 454 Preservation of Stone , 455 INDEX OF PIOMENTS. yro Jind a pigment^ and the reference number to the page where it is treated o/^ look first for the proper color ^ and under that heading find the trade name as it appears in alphabetical order. ''''See''' indicates a synonymous najne., or a specific name referred to its general class.Ji Artist's Pigments 361-364 Car Pigments 361 Carriage ^ ' 360 Preparation of 364, 365, 437-439 Decoration, Pigments for use in 361 Drop Lakes 360 Pigments changed by Arsenic 372 Iron 328, 372 Chalk 370, 371 " " Lead 369, 370 " Lime 370, 371 not affected by Lime 371, 372 changed by Sulphureted Hydrogen 369 " White Lead 226, 370, 376 " Position in the Normal Spectrum 306 Sunlight and Pigments 374-381 Protecting Colors 379 Testing Pigments 33-48, 293, 295, 296, 379 Water Colors — Poisonous 373 Washing Pigments 294, 366 BLACK PIGMENTS. Almond Black 298, 300 Berlin Blue (see Prussian Black) 298 Black Lake 300 Black Ochre 298 Bone Black 297, 299, 376 Charcoal 299 Coffee ■ 298 Drop 299, 376 Cork 298 Frankfort " 298, 300 Graphite 300 Ivory " ... 297,298,299,376 Lamp 297, 298, 300, 301 xii Painting and Faintei's' Materials. Manganese Black 298 Paper 298 Peach stone 298, 300 Prussian 298 Prussian " (Ochre) 298 Purple (from Madder) 298 Spanish 298 Swedish See Charcoal Black. Vine 298, 299 BLUE PIGMENTS, Antwerp Blue 338, 339 Azure See Cerulean. Berlin 306, 338, 355, 369, 377 Bice (see Copper Blue) 340 Blue Ashes 340 Blue Verditer. See Copper Blue. Bremer Blue 341 Cerulean 337, 351 Cindres Blues. See Copper Blue. Cobalt Blue 306, 336, 351 Copper artificial 3400410^9 " " natural 340, 369 Cyanine 337 Hamburg'' (see Berlin Blue) 338 Indigo 339, 340, 335, 377 '' Carmine 340, 351 Intense Blue 340 Lapis Lazuli. See Ultramarine, Natural. Lime Blue 341 Manganese Blue 341 Mineral '' See Berlin Blue. Newburg 339 New Mineral Blue. See Newburg Blue. Night Blue... . 339, 351 Paris " 339. 347, 351, 361, 369 Prussian 339, 347, 351, 361, 369 Saunder's See Cyanine. Sky • 337 Smalt ... 336, 335 Thenard's Blue (see Cobalt Blue) 337, 351 Turnbull's 339 Ultramarine, Artificial 306, 333-336 " French. Sec Artificial. Green 333 Natural 306, 332, 333 " Violet 336 Index to Pigments, xiii BROWN. Asphalt 147, 148, 358, 359 Berlin Brown 359 Bistre 358 Mineral Bistre 358 Brown Ochre 42, 357 Cappagh Brown 358 Cassel Brown 358 Chrome Copper Brown 359 Chrome Iron Brown 359 Cologne Earth 358 Madder Brown 358 Mahogany Brown 357 Manganese Brown. See Mineral Bistre. Mixed Brown 351-357, 359, Appendix Mummie 359 Sepia 358 Terre de Sienna 43, 356, 357, 366 Burnt..., 43, 357 Sienna, American 43 Spanish Brown 358 Umber 43, 357 Burnt Umber , 44, 357 Italian Umber 357 Vandyke Brown 44, 358 Velvet Brown 357 GRAY. Silver Gray 360 Zinc Blende « 355 Zinc Gray. See Zinc Blende. GREEN PIGMENTS. Alexander Green. See Veronese Green. Bremer 346 Bronze " 347, 361, Appendix Brunswick " American 346 not Poisonous 346 Poisonous 346 Chrome Greens, Genuine 343 Chrome (Yellow) Greens 347-350, 369, 377 Cobalt Green. See Green Smalt. Emerald 306, 345, 346 Eisner's 343 Green Earth 343 Green Lake 346 Greens by Black 350 for Artificial Light 3^1 xiv Painting and Painters' Materials. Green Smalt 343 ^* Vermilion 342, 349 Guignet's Green 342 Hooker's 350, 378 King's 346 Leaf Green. See Chrome (Yellow) Greens. Leipsic 346 Malachite Green. See Veronese Green. Manganese , . . 347 Mixed Greens 347, 351, Appendix Moss Green. See Chrome (Yellow) Green. Mountain Green (see Veronese Green) 344 Munich 346 Napoleon's See Green Earth. New Green 346 Night Greens 351 Oil Color Green. See Veronese Green. Oil Green. See Chrome (Yellow) Green. Paris Green 345, 346, 370, 380 Patent 346 Permanent Green ^ 342 Pickle 346 Prussian " 350 Rinman's See Green Smalt. Sap " 346 Scheele's Silk See Chrome (Yellow) Green. Stannate of Copper 343 Stone Green. See Veronese Green. Swedish 346 Sweinfurth's Green 346 Swiss 346 Terre Sienna Greens 349 Terre Vert 343 Turk's Green 343 Tyrolese See Green Earth. Unchangeable Green 351 Verdigris 344, Appendix Vermilion (Green) 349 Veronese ^'^ 344 French 344 Vert Pellitier. See Permanent Gr^en. Vert de terre. See Green Earth. Victoria Green 342 Viridian. See Guignet's Green. Vienna Green 34(5 Zinc Yellow Greens 348 Index to Pigments. xv RED PIGMENTS. Antimony Red 304, 319 Orange 319 Brazil Wood Lake. See Red Wood Lakes. Burnt Carmine Lake 313 Burnt Red Ochre 43, 305, 321 Cadmium Red 304 Carmine Lakes 304, 306, 312-314, 378 Chinese Lake. See Carmine Lake. Cochineal Lake 312 Carmine Munich Lake 313 Coal Tar Colors 297, 308, 305, 314 Coraline Lake 305 Chrome Red 304, 314, 316, 323 Crimson Lake 304, 312 Eosine Lake 304 Eosine Vermilion. 315, 316 Field's Madder Carmine 309 Florentine Lake. See Carmine Lake. " (Spurious.) See Redwood Lakes. Fuchsine Lake 314 Fuchsine and Pernambuco Lake 305, 314 Hamburg^ Lake (see Red Wood Lakes) 304 Indian Red 305, 321 Iron Minium (see Metallic [Iron] Paints) 179 Iron Oxide 128, 129, 305, 320, 321 King's Red 304 Lac Lake 304 Light Red Ochre 304, 321 Madder Lakes 304-309, 378 " Carmine Lake , 304, 309 " Artificial 308 Mars Colors 305, 321 Metallic (Iron) Paints 37, 40, 41, 90, 114, 142-146, 155, 156, 161, 179, 414 Munich Lake, Genuine 304, 309 Pernambuco Lake. See Redwood Lakes. Ponceau 304 Purple Lake 304, 309 Red Lead 41, 89, 90, 94, 113, 115, 142-146, 169, 179, 198, 200, 202, 203, 2i3» 304,319, 369 Redwood Lakes 311, 312, 370 Roman Lake (see Redwood Lakes) 304 Rose Carmine Lake 304, 309 Rose Pink. See Redwood Lakes. Scarlet Lake 304 Scarlet Ochre. See Light Red Ochre. Tuscan Red 305, 321 X7'i- Pamtmg and Painters^ Materials, Vandyke Red 305,321 Venetian Lake (see Redwood Lakes) 304 Venetian Red 180, 305, 320, 414 Vermilion, Rules for Use of 317, 318, 369 American 304, 315, 369 " Carmine 304 Chinese 304, 317 European 304, 306 English 304,^317 Substitute 304, 317 " Iodine 374 Orange 304, 319 Vienna Lake. See Redwood Lakes. Zubia Red. See Vermilion Substitute. YELLOW PIGMENTS. Aureolin 322, 327, 328 Baryta Yellow (see Lemon Yellow) 322 Cadmium Yellow, deep 322, 326, 327, 369, 370, 371 lemon 322,326,327,369,370,371 pale 322, 326, 327, 369, 370, 371 Chinese 330 Chroma " 306, 322, 323-325, 369, 371 Orange Chrome Yellow 322, 323 Cologne 322, 325 Dutch Pink 322 Gamboge 33^1 35Si 3^9 Italian Pink. See Dutch Pink. Indian Yellow 322, 323 Iron 330 Jaune Indienne. See Iron Yellow. King's Yellow 330 Lemon Cadmium 326, 372 Lemon Yellow -^zo.^ 327 Litharge 89, 90, 169, 202, 203, 331 Mars Yellow 322, 330 Mineral See Chrome Yellow. Mosaic Gold 330 Naples Yellow 322,323, 328, 369, 370 Ochres Yellow 42, 294, 295, 322, 329, 330, 366 Bermuda Ochres 329 Gold Ochre 330, 362 French Ochres 42, 329, 330 Roman Ochre . . 329, 330 Stone 329, 330 Orange Mineral 331 Orpigment 330 Index to Pigments. xvii Quercitron Lake 322,323,351 Raw Sienna 322, 323, 366 Strontian Yellow 322 Turner's Yellow 330 Yellow Lake 322,323, 378 Yellow Madder 330 Yellow Mineral 322 Yellow Ultramarine 322, 325 Zinc Yellow. See Yellow Ultramarine. WHITE PIGMENTS. Antimony White.. « 302,369 Arsenic 302 Barium See Barytes, Artificial. Barytes, Natural 34. 38, 39, 40 Barytes, Artificial 39, 302 Barytic White. See Barytes, Artificial. Bismuth 302, 369 Blanc d' Argent 301 Blanc Fixe. See Barytes, Artificial. Cadmium White 302 Carbonate of Lime (Chalk) 302 Carbonic Acid Lead 116, 117, 122 Ceruse (see French White Lead) 301 Charlton White 302 Chinese 301 Constant See Barytes, Artificial. Cremnitz 302 Flake , . 226, 301 Fulton " 39, 302 Gypsum 302 Hamburg White, Venetian White, Dutch White, found only in books. Hydrated Oxide of Lead 36, 116, 117 Krems or Crems White 301 Kremser iii, 236 Lime 34, 302 Lithopone. See Patent Zinc White. London White 301 Nottingham White 301 Oxychloride of Lead. See Pattison's White. Patent Zinc White 39, 180, 302, 303 Pattison's White Lead 112, 301 Permanent See Barytes, Artificial. Silver 301 Strontian 302 Sulphate of Lead 301, 414 " ^' Sublimed (White Lead) 301,414 xviii Painting and Painters' Materials. White Lead 34, 35, 36, 44-49, 68,108-123, 151-155, 183, 187, 236, 259, 301, 368, 369, 370, 374, 413, 414 " Dutch 108, 123 English 413, 414 French 108, m, 120 German 108, 110, 301 Gruneberg's 112, 200 Sublimed 301, 414 Zinc Sulphide. See Patent Zinc White. " White 37, 45, 46, 68, 90, 94, 179, 180, 199, 236, 296, 297, 301 PART I. * * * It seems to me most essential to call the attention of those engaged in the business of coach painting to the great advantage, if not the actual necessity, of acquiring something more than a mere superficial knowledge of the materials and compounds continually passing through their hands, and of the conversion and application of such mate- rials which play so important a part in their daily vocation."^ — THE HUB'S Prize Essay on Carriage Painting. CHAPTER I. Things in General. Considered as a useful art, painting is a process of water-proofing. It is its special service in construc- tion to prevent wood and iron from becoming moist. The painter's first duty is, therefore, to understand the action of water upon these substances, and es- pecially how to prevent it. 1. The action of air alone upon seasoned wood, even if the air contains considerable moisture, does not cause it to rot. The woodwork of ancient churches and cathedrals in Europe has remained sound, although exposed for centuries to air and without paint. Tredgold says, " For timber that is not exposed to the weather, the utility of paint is somewhat doubtful." 2. The action of air containing not more than the usual moisture of air in our climate is to swell and split wood rather than to rot it. In preventing this action of moist air, therefore, we need to treat all sides of the wood facing air. These are practically all sides of the wood not in immediate contact with some solid substance. 3. The first cause of decay in wood is the presence of sap. Painting on such wood only prevents the water of the sap from escaping, and in the confined air the fermenting sap rots the wood the faster. The water also in seeking to escape may raise blisters upon the paint, as will be discovered by picking the blisters, when it will run out. It is probable also that 2 Painting a7id Painters' Materials, seemingly empty blisters are formed by the gases of such painted wood. 4. The second and most active cause of the de- struction of wood is dry rot. This comes with the growth of little plants (fungi) which feed on the woody fibre. They thrive best in warm, damp and close places, as under the floors of warm cellars or basements, but once fairly started are able to live even in a current of not too cold air. Timber built into walls, or over walls not yet dry, is specially sub- ject to it. Painting over unseasoned timber is also favorable to it — as in fact anything is which shuts out the air from moist but not too damp wood. A piece of tim- ber apparently sound may be full of dry rot, as may be discovered by boring into it. Wet rot is the general name given to the rotting of wood due to moisture, but not to the presence of the dry-rot fungus. The painter has more to do with this than with dry rot. He should strongly advise that no two pieces of wood should be put together until their surfaces have been covered with paint, un- less they are to be protected by their position from all access of water. In his own special work the painter will see that all cracks are filled with oil and white lead or good oil putty, and that they are thoroughly filled, not merely skimmed over. 5. The alternate wetting and drying of wood causes decay, but of uncertain rapidity on surfaces exposed to free sunlight and air. The painter is more con- cerned with the checking, swelling and warping caused by such exposure, and also by that roughen- Rusting of Iron. 3 ing and breaking up of fibres which unfits the surface to be a foundation for fine painting. (a) To sum up the matter in a word : all wood con- taining or exposed to moisture {and not submerged in water or earth) needs treatment in proportion as it is shut out from free access i:) the air, (b) On the other hand, to shut out air from wood is to hasten its decay unless the wood remains quite dry. (c) Under the ordinary division of labor among the trades it falls to the painter mainly to prevent the warping, swelling and checking of wood, and the con- sequent opening of joints and cracks. We shall return to this subject. RUSTING OF IRON. Iron will not rust in perfectly dry air; the presence of some water is necessary, and it rusts more rapidly in warm countries than in cold countries and in warm than in cold weather. Cast iron rusts least rapidly, wrought iron perhaps 25 per cent, and steel about one-third faster.* The presence of acids (as car- bonic acid) is especially active in causing rust. Rust contains water, and by its own action causes rust. It is necessary, therefore, to prevent its begin- ning at any point, or it may extend rapidly over the entire surface, and even under those portions which have been painted. Galvanized iron is protected from rust only so long as the entire surface is perfect ; if broken at any point, the action between the two metals may soon destroy the covering of zinc. * It is doubtful whether steel rusts more rapidly. It probably does not if the black scale (black oxide) is removed. Steel should, perhaps, be said to rust more irregularly rather than more rapidly than iron. 4 Painting and Painters^ Materials. The black (scale) oxide of iron which forms on heated iron (but is apt to scale off rapidly and easily) is a protection to iron against rust. In painting iron it needs to be removed because it so readily becomes loose, carrying the paint with it— from steel because it produces pitting." The black magnetic oxide of iron (produced by ex- posing iron to a temperature of 500 to 1,200 degrees in a chamber containing superheated steam) is prob- ably its best protection against rust, as this oxide re- mains attached to the iron, and if detached at any point by violence the rust does not extend from that place to the others covered by the oxide, as it might do in the case of paint or zinc-covered iron. This method is known as the Bower-Barff process. Very thin plates of iron rust more rapidly than thick plates, because the scales of rust are thrown off by the greater expansion and contraction of such plates, exposing, as they do, so much surface to heat and cold. To sum the difficulties with iron: (i.) Iron rusts in the presence of even moist air. (2.) Although the black (oxide) rust protects the iron, it easily scales, while the red rust itself produces rust, which is not the case with substances forming under like cir- cumstances (to a degree) on zinc and copper. (3.) Paint has little to hold by on smooth iron ; and (4), the rust can spread under it from any rusted uncov- ered point. (5.) Finally, the expansion and contrac- tion of iron under heat and cold tend to crack the un- elastic layers of paint. Frequent inspection, brushing away the points of rust and retouching the bare iron, and finally frequent Rusting of Iron, 5 repainting of the entire surface, form the only sure protection to the surface of painted iron. For further facts about iron see Rust 2.xi^ Painting Iron, CHAPTER II. The Cracking of Varnish and Paint. Varnish and paint disappear in three ways (i) by cracking, (2) by peeling, (3) by perishing. These are simple processes ; it is quite easy to assign causes for them ; but this, as we have found, is not quite the same as discovering what the true and most active causes of destruction are. The principal agents in destruction are undoubtedly the sun, water, cold, the oxygen of the air and the friction of storms of wind and rain. Paint not ex- posed to the sun, to violence or to moisture will last a very long time. Paint is of the same nature, whether it is on a canvas or a carriage ; and the study of facts about pictures will give us much lio^ht about carriages. Dr. Liebreich says : " The amount of exteraal injury oil paintings sometimes endure and stand is perfectly amazing. Pictures in the course of cen- turies, during the destructive fury of wars and revolutions, may have been torn out of their frames, rescued from below the ruins of burned monasteries, may subsequently have passed from one bric-a-brac shop to another, where they have been piled up to be pulled about at each new inspection, and literally trodden under foot, whereby they have finally been reduced to a state of color- less, grayish, or black rags. Still such pictures may not unfre- quently be awakened, as it were, to new life, to their original bril- liancy of color, if, with all necessary care, their injured limbs are put together again, their wounds are healed, and fresh nourish- ment, air and thorough cleansing are administered to their lacer- ated bodies. ' ' A sound constitution is, of course, a necessary condition for obtaining any such result ; without it we can only obtain a partial cure." The Decay of Picture Varnish. 7 Prof. Max Pettenkoffer, one of the ablest scientific men in Europe, has, fortunately, investigated the de- cay of varnish and paint on pictures. Many old pic- tures become invisible by a change in the varnish, formerly supposed to be due to mold. This was proved to be a mistake, and the true cause of the whitening and lack of transparency shown to be par- tial separation of the gum of the varnish from the hard, dried oil. A piece of window-glass becomes white and no longer transparent when ground to a powder, without any change in the substance of the glass — merely a change in its form. Pettenkoffer proved that this powdering of the glassy varnish was due to the evaporation of water which had been deposited by the air upon the surface of the picture ; and he produced a like result by frequently wetting the surface of a picture, allowing the water to evapo- rate as if put there by natural means. The immediate cause of the whitening is, as we have said, the result- ing separation of the varnish gum from the oil in which it is held.* In the great gallery at Munich, 52 per cent, of the pictures in rooms with a northern exposure were found to be more or less affected by this powdering of the varnish ; in rooms with a southern exposure but 10 per cent, were so affected. We may, there- fore, rest quite certainly on the fact that the evapo- ration of moisture has much to do with the decay of varnish. The influence of moisture is the more evi- dent from the fact that other galleries than those at Munich suffer less from moisture and more from * For proof and further explanations see Changes in Varnish^ under Carriage Painting. 8 Painting and Painters' Materials, other causes. The dampness of these galleries com- pared with the dry and brighter warmth of those in Italy shows that only the direct influence of the sun is injurious to paint ; to direct sunlight and the rain, hail and snow, we must look for the destructive agencies which lie outside the paint itself. CRACKING. Intending to return to the consideration of perish- ing, we may here profitably study the cracking of paint and varnish. One turns naturally to contrac- tion and expansion of wood as a cause of this crack- ing upon all wooden structures. The following table, the result of French experi- ments, is interesting in this respect : Increase of Measurement when Thoroughly Saturated with Water ^ The Increase Represents Proportion to loo parts of Measure. Length. Breadth. Maple 0.072 3.350 Apple 0.109 3.000 Birch 0.222 3.860 Pear 0.228 3.940 White beech 0.400 6.660 Purple 0.200 5.030 Box 0.026 6.020 Cedar 0.017 i-SOo Ebony 0.010 2.130 Young oak 0.400 3.900 Old oak 0.130 3.130 Young ash 0.821 4.050 Old ash 0.187 3.840 Fir-wood 0.076 2.410 Perhaps more satisfactory measurements have been made in Germany by Karmarsch, which are so valua- ble that we give them also in full. The Contraction of Wood, p Amount of Shrinking of Green Wood i?t Percentages. Cross Section. Kind of Wood. In length.* Directly In direction t through, f of yearly rings. 0.062 to 0.200 ... 2.0 to 5.4 4.13 to 7.3 2.228 2.9 to 3.94 5.5 to 12.7 0.20 to 0.34 2.3 to 6.0 5.0 to 10.4 0.028 to 0.435 I.I to 7.5 2.5 to 10.6 0.187 too.821 0.5 to 7.8 2.6 to II. 8 0.076 I.I to 2.8 2.0 to 7.3 0.008 to 0.201 0.6 to 3.8 2.0 to 6.8 0.013 too.288 0.3 to 7.3 1.4 to 7.1 0. no 1.09 1.79 0.223 2.6 to 8.2 4.0 to 17.6 0.068 to 0.62 1.2 to 4.2 2.8 to 9.8 0.086 tOO.I22 1.7 to 4.82 4.1 to 8.13 0.014 to 0.63 1.2 to 4.6 2.7 to 8.5 * In line a to a. About i-io of i per cent. + In line of layers marked c. About 5 per cent, $ In line of z to /. About 10 per cent. lo Paifiting and Painter s Materials. It will be seen that the expansion of wood due to moisture is from 5 times (in young ash) to 213 times (in ebony) greater in its breadth than in its length. One would naturally expect, therefore, to find cracks in paint and varnish parting as the fibres of the wood part — namely, parallel with its grain. The facts are that those cracks, extending with the grain, are much less numerous and of less importance than those extending across the grain. Within the house, or car, nearly all cracks run at right angles to the grain of the wood. Out of doors, except with very green wood or very poor materials, cracks of much consequence do not appear in the wood until this has become dry and shrunken, and has expanded again in some spell " of bad weather, or by the ad- mission of water through cracks in paint or varnish, which cracks first appeared across its grain. It is surprising how much a door or panel may shrink without cracking its paint — at least without cracking it as wood always checks," namely, with its grain. The reason is very simple. In the shrink- ing of wood, every little fibre comes closer to its neighbors, and the larger fibres, especially, raise the paint or varnish in little folds, which remain as so much " slack," ready to be stretched again. The strain at every point, whether from contrac- tion or expansion, is thus made very gentle, as one may see from the fact that a deep " check " in the wood cracks varnish only when and at that place where the wood itself is torn apart. Previous to this the total strain in the wood must have been very considerable, but it did not crack the varnish. The contraction of wood in breadth, and its small The Effect of Heat on Varnish. ii contraction in length, may, however, give direction to the cracking of paint and varnish. If the contraction of the paint or varnish upon itself produces the cracks, then the greater contraction of the wood in breadth will prevent cracks extending with the grain of the wood. Even this, however, does not seem to be the principal reason why varnish and paint crack at right angles to the wood fibres. Does Heat Crack Varnish ? — In order to make sure that the contraction of wood has little effect on good paint or varnish, one may observe the influence of heat on painted surfaces. Heat does not crack paint or varnish; in constant heat both are made elastic. The cabs on the loco- motives of a railroad afford an excellent illustration. These locomotives have unusual care, and are well varnished. Nearly all outside and inside surfaces of the cabs are, nevertheless, cracked across the grain of the wood. The exception is the ceiling of the cab, which is at times so hot the hand can hardly be al- lowed to remain upon it. The wood is badly shrunken, but the varnish is fine and glossy, without apparent cracks, except a few above the side windows, and these are across the grain. The panel back of the steam-gauge is also free from cracks, while the sash of the small windows on either side are full of fine cross-grained cracks. The panel is thick and pro- tected; the sashes are thin and exposed to the cold air, for it is the cooling which cracks varnish. Var- nish and paint crack, as a rule, from contraction in their own substance. The effect of heat is to expand and soften varnish. Heat also unites the separated varnish, as a 12 Painting and Painters' Materials, clouded and whitened varnish surface can be made clear by heating. Let us now examine the positive side of this cracking, ist. On a perfect surface paint and varnish have no special direction of cracking, but may part in all directions. On leather, or any pressed or stretched surface, cracking quite often takes a circular form, one circle outside another, 2d. Cracking will occur at that point where the strain is greatest and there is the least elasticity. 3d. The cracking of paint and varnish on wood is controlled commonly as to its direction by the fibres of the wood. If we place together two pieces of wood with smooth surfaces, it will be found that they can be rubbed much more easily with than across the grain. Across the grain each little fibre is meeting resistance from each fibre on the other piece, espe- cially from those larger ones which mark the close of each year's ring or growth. If we place a drop of oil upon a smooth piece of wood, and let fall gently into it other drops, the oil will be seen to move fur- ther with than across the grain. Each added drop will make this fact plainer, and if a glass " dropper ' be used, the difference between the results of pointing the force of the drop with the grain, or across the grain, will make the fact still more marked that all movement on the surface of wood is easier with the fibres than at right angles to them.- In all contraction of the paint upon itself, there- fore, the movement (contraction is movement) will be easier with the grain than across the grain. This will tend to open cracks across the grain. An Experi77ient. — For a test, Messrs. G. W. Read The Cracking of Varnish and Paint, ij & Co. sawed for us from the damp log, one piece of whitewood, and several of black walnut. These were planed and varnished with four or five coats of com- mon rubbing varnish, one over the other in quick succession. After a few days (meanwhile remaining in cold room) they were placed under a powerful heat for about fourteen hours of each day for a week, the unvarnished and then the varnished side being ex- posed. The wood was by this time very dry, but no cracks appeared in the varnish, although the white- wood had become badly warped. The whitewood was then soaked in water until it had warped in the opposite direction to the shape of a bow, when it was again exposed to heat until dry. The soaking pro- duced some cracks, with the grain, and a few cracks appeared, caused evidently by this second baking, as a few indicated by their shape that they were con- trolled by the warping. But the cracks were hardly cracks at all, and it would now, many months after, be difficult to find them without a glass. This experiment shows very plainly, that at least until it loses its elasticity wood by contraction does not crack paint or varnish. Cracks with the Grain. — Many cracks extending with the grain of wood are no doubt due to the ex- pansion of the wood by moisture. They point to the fact that the wood has not been properly protected from moisture. Even if painted on one side, wood in the interior of the house gains from 3 to 6 per cent, of its weight of water in summer, and loses it again in winter.* The gain and loss of wood ex- * This fact I have through the kindness of Professor Brewer, of Yale College, who averaged for me the result of very extended but still incom- plete experiments. The figures are true of small not very thick pieces. 14 Painting and Painters' Materials. posed to the weather must be much greater than this. We cannot entirely prevent the action of moist air on wood, but this is not necessary. Good paint and var- nish properly used will not crack on this account if the wood is reasonably protected. I speak here of severe cracking ; small cracks, of course, will appear, but these are of little consequence. They are easily covered, as are any spots of perishing." The fatal disease in paint and varnish is severe cracking. Cracks Across the Grain, — The best examples of cracks across the grain of the wood are those to be found on carriages. A carriage spoke almost inva- riably cracks in a circle, or some arc of a circle, across the grain of the wood. A jar will crack varnish more or less across the grain of its wood, as may be seen on that part of the carriage to which the mounting steps are attached. The jar on the wheels undoubtedly does aid in the cracking. I have made one experiment with a com- paratively crackless spoke, and have produced cracks by striking it upon its end. These cracks have a ten- dency to run diagonally across the grain, some also directly across. On some spokes, at least, it will be observed that the larger cracks are half circles in the central part of the body of the spoke, extending from and to, but not crossing, a line on each side drawn from each side the tenon, and are much more fre- quent near the hub. But notwithstanding all these appearances, the facts appear to be that even on spokes it is not the strain and jar which cause these cracks across the grain. Out of 50 show spokes hangmg in a paint shop 48 were cracked in more or les.s complete circles around The Cracking of Varnish and Faint, 75 the spoke. The reason why varnish on spokes has more tendency to crack near the hub is because there is more varnish left there. Cracking on Carriages, — Cracking first appears on the felly or rim of the wheel, and is partly caused by the strain, and partly due probably to the effect of mud and other substances on the paint and varnish. The cracks almost invariably extend across the grain, because the strain will stretch the varnish in this direction in much the same manner, as though the felly were to be straightened to a flat piece. On the body the cracks are generally across the grain ; but not always, because if the cracking does not begin in coats next the wood, but in the varnish coats, the wood fibres have no influence on the direc- tion of the cracks, because they are completely cov- ered by the filling coats. On the top of coaches, the cracking is often caused by a change in the muslin, and may often be the fault of the method of cover- ing. It is also caused by the amount of white lead, etc., which is placed under the canvas. It is prob- ably better to glue on the muslin. It may also be caused by the white lead over the canvas, if it has not been properly dried or contains too much oil. The roof panels may also cause cracking ; but as the mus- lin, etc., lies between it and the varnish, the fibres of the wood can have no influence over the direction of the cracks. On light wagons, cracks under the seat-board are caused by strain. On carriages and coaches exposed to a rain (for such it seems to be) of ammonia in stables, cracks will be found on upper and less cov- ered portions, while none may exist where the varnish i6 Painting and Painters Materials, has been protected from the shower of ammonia. Usually, those parts of a carriage least exposed to (i) street mud and (2) sun, are not cracked, or if they are, they indicate a poor job, hence these portions are the better places to observe as test. Other parts of the carriage may have had good work, but hard usage or poor care — as, for example, frequent washings with poor (/. alkaline) soap. Cracking on Cars, — On cars with good surface, small cracks will run in all directions. The larger cracks are generally across the grain of the wood, be- ginning often at the edges of the panel near panel strips, because at this point a greater thickness of paint and varnish is found. Occasionally one finds a car with long, deep cracks extending with the grain of the wood. It is possible that in these cases the wood may be at fault, but if so it shares the blame with the rubbing varnish. But even if the wood be at fault in these cases, it is the expansion and not the contrac- tion which causes the cracking. I cannot prove this, but think it might be proved by careful observation and experiment. Within the car nearly all cracking extends across the grain of the wood. Fewer cracks will be found on veneered surfaces, and these not always across the grain, probably because the shrinking of the wood on which the veneer is placed has some influence. Interior of Buildings. — Cracking is usually across the grain, except about window sills and on doors ex- posed on one side to damp air. Exterior of Buildings. — Failure to treat more than one side of the wood used in building (clapboards, etc.) shows very plainly in cracking of paint on build- Cracking of Paint and Varnish. ings, which is often with the grain of the wood. These cracks are not entirely caused by the wood, but proper treatment of all its sides would better pre- serve the paint. Cracks on Furniture. — The cracking of varnish upon furniture is quite uniformly across the wood fibres, except in the case of veneers, which occasion- ally offer an exception, owing partly to the peculiar way in which they are cut, and partly perhaps to the effect of the wood on which they are fastened. Old furniture, not cracked, if examined with a magnifying glass and the light reflected from the surface of the varnish caught with the glass, will be found to have numerous straight lines running across the grain. These are probably due to the parting a thin layer of the surface varnish which has lost its substance by slow decay faster than the coats beneath it. To sum up the decay of varnish : I St. Properly treated wood has little destructive influence on paint or varnish. 2d. The so-called "perishing" of paint or varnish- ing is of less consequence, except as a preliminary cause of cracking. It is easily cured by other coats. 3d. The great and fatal diseases of paint and var- nish are cracking and peeling. 4th. Varnish very largely (and paint to a degree) furnishes the force which cracks and destroys it. 5th. If these conclusions are trustworthy, we have these practical rules : (a). Cracking of varnish or paint upon wood in a direction across its grain indi- cates poor material or an improper use of good ones; /. the coats not being old nor exposed for many hours per day to direct sunlight, and sufficient time i8 Painting and Painters' Materials. (which is not often the case) having been allowed for the painting. The painter's most practical question is, how safely to save time. (b) . If the direction of the cracking is with that of the wood-fibres, the gain of moisture by the wood is to be suspected. The wood has not been properly protected on some or all sides. (c) . Any strain, however brought on the varnish, tends to open cracks across the grain because it is more bound in this direction, and moves more easily in the other — /. with the grain. CHAPTER III. The Purchase and Manufacture of Varnish. As a rule, with few exceptions, varnishes wilt be purchased; information as to manufacture may, there- fore, be limited to a purchaser's needs, and to some notes on the simpler varnishes. There is more difficulty than mystery about the manufacture of varnish, and this difficulty arises from lack of means of testing the materials used, the heat of the fire, etc. Varnish is made of oil, turpentine and resins. The quality and manipulation of each of these materials determine the quality of the varnish. Varnish makers are perhaps inclined to feel that other people know very little about varnish, and other people are perhaps quite as likely to get their heads full of notions as of hard facts about varnish. On the other hand, varnish makers know much less than they should. The whole subject is in that practical" stage of knowledge where a notion is in danger of getting more valued by everybody than is a scientific fact. The facts see7n to be: i. The hardness and part of the durability of the varnish depend upon the resin. The less water the resin already contains, the less it is affected by the air; and for this and other reasons, the less by water. The air oxidizes the resin as it does the oil, but hard resin protects the oil from the air. Water separates the fine particles of resin from the oil. 20 Paintmg and Painters Materials. 2. The oil gives the varnish its elasticity; if it has been made into soap or contains much drier, the elasticity of th% oil is destroyed. Better for durabili- ty a poorer, resin and more good varnish oil than a better resin and poor or little oil. On the other hand, the more oil in the varnish the more skill needed in using it coat over coat. Also the more skill needed in making it a varnish. In selecting a varnish, as we shall see, three facts cannot be left out of account: i. Who is to use it? 2. What is to be put under or over it, now and by and by ? 3. When jviH it be used, and how much time will it have to dry ? The Resins, — There are in all about thirty differ- ent resins used in varnish making — different in name, and many possessing peculiar qualities. The best of these resins are fossils, that is to say, they are dug out of the ground or are found in the beds of rivers where they have lain for unknown years. All of the fossil resins are from trees, which may for practical purposes be called species of pines. The hardness of resin appears to depend upon its age and the amount of pressure it has undergone m the earth. See Theory of Linseed Oil, Amber. — This, the hardest of all resins, is found in layers or mines on the coast of Prussia. Only ref- use, or so-called ''black'' amber, is used for varnish. The trees from which came the amber grew, perhaps, when no men existed on this earth. *' Amber" var- nish usually means merely amber-colored varnish. CopaL — Next in hardness to amber is the best copal from Zanzibar (Africa), called by the English animi. It is very difficult to dissolve, and practically The Manufactitre of Varnish, 21 there is but one method in use — namely, distiUing the resin until it loses from 20 to 25 per cent, in weight. It may then be dissolved in boiling oil. Zanzibar copals are of three sorts in many grades; the better sorts give most lasting varnishes, but with more ten- dency to crack. Sierra Leone Copal is much used in English var- nishes. It has nothing to do with Sierra Leone, but comes from the river beds of the interior. It is the only African copal which is dissolved in cold alcohol Its color is not always so good as Zanzibar, or the best Kauri copal, but it is harden than the latter. In English varnishes it is mixed with Zanzibar (or "animi"), the Sierra Leone giving elasticity, the animi hardness. There are also of African copals, pebble, ball or glass, Accra, Loango, Gaboon, Congo, and the three sorts of Angola and one of Benguela. The pebble, ball, best Loango, red Angola and the Benguela are much esteemed — the white Angola being the softest its softer kind sticking to the teeth when chewed. The pebble (or pebblestone) is the hardest of these copals. The Kauri or Cowrie, of New Zealand, is, how- ever, the principal copal used in America — its com- sumption being probably ten times greater than the combined quantities of those already mentioned. It is from two to nine times cheaper in price, is more colorless, very easily dissolved, and gives, therefore, a clearer varnish than any other copal resin. It is allied in chemical composition to the dammar resin; it m^lts more easily than mastic, but less easily than rosin. Its great defect is its quick loss of lustre — the loss 22 Painting and Painters' Materials, of the prime requisite of a varnish. A varnish made from the best selected pieces of Kauri gum holds its lustre second to a mixed Sierra Leone and Zanzibar copal varnish, but lower qualities go quickly. The loss of lustre does not cause loss of protecting power; the substance of the varnish remains. Anime . — This was the original name given to copal, from the fact that it contained insects embedded in the gum. It is now the technical name for South American copal. This copal is from Brazil, and is not found in our market. Manilla, — There -are two sorts (neither fossils), a harder and a softer, from the Phillipine and other islands, Borneo, Singapore, etc. They find a large market here, and are used in varnishes alone and with harder resins. Rosin is too well known to need many words. By it, however, we may illustrate the making of resin in the tree. All resins are dissolved in their natural turpentine; they are entirely different from muci- lage. Mucilage is a cell wall (see cut) dissolved by water: the mucilage of the linseed, for example, which soaks up the saliva from the mouth, swells up and changes into a sweet slime. A resin is a secretion or perhaps a waste-stuff (excretion) of a tree or plant. The resin which we now use for varnishes flowed out ages ago in a fluid state, often catching and holding the feet and then the bodies of insects light- ing upon it. By various causes, gradual or sudden, it found its way underground. There it was pressed upon by the earth. To the action of the oxygen of the air and the pressure of the earth, its hardness is The Manufacture of Varnish, 23 due. The harder resins contain least water; dammar, which is a recent resin, the most water. The following are the average prices and the order of quantities of the different resins, as used in Ameri- can varnishes : Order by quantity. Prices per lb. 1. Kauri 10 to 50 cts. 2. Manilla 10 to 25 3. Dammar 16 to 25 4. Zanzibar, best $1 to $1.25 5. Benguela 85 Section through a resin passage of abtes excelsa (fir and spruce trees). The cavity H as well as the thin walled cells H are filled with semi-fluid resin. The thick wall cells P contain starch. English Varnishes. — The great market of the world for resins, as well as for money, is London. English varnish-makers have, therefore, the first opportunity, and English varnishes have always been the best in the world, /. the best varnishes (as we used to say in the old anti-slavery d^ys) per se. 24 Painting and Painters' Materials, American Varnishes, — The poorer sorts contain rosin and many other things ; the quality of the bet- ter sort we have discussed below. Qualities of Varnish. — The following may be given as the qualities of a good varnish : 1. It dries within 30 hours so as no longer to be affected by dust, and presents after three days a bril- liant, firm surface with no spots of dull or fatty as- pect. 2. It retains for at least 9 or 12 months a clear, elastic and transparent surface. Color is not important, unless it refuses to bleach to colorlessness in a thin unstopped vial exposed to sunlight for three to five days. The durability of varnish must be tested by trial. Prime a smooth board with white lead and oil, let it dry, then give one coat of varnish or one coat of several kinds of varnish in squares. Expose to the weather, and make observations from time to time during one year. Don't trust yourself to make this test unless you are an honest man, and do not use samples" of varnish for this purpose. Samples sometimes are less durable than a regular article drawn from a barrel, sometimes they have more dura- bility. It is not advisable, therefore, to trust samples. It is also much better to prepare three boards than one, as slight differences are due to position ; for ex- ample, either end of an upright board will give a severer test than the centre. Such a test, honestly made, is fair, and, if dupli- cated and repeated, quite satisfactory. The question arises, however, what is to be the standard of judgment. A kauri g\xm loses its lustre Qualities of Varnish, much sooner than a true copal gum, although the best grade of kauri follows well after copal. The loss in lustre is not, however, loss in protecting power either from water or dust. It is, however, loss of the prin- cipal quality for which varnish is used, namely, reflec- tions of white light. It also may show the gum to have lost its power of clearing itself of water from its surface, and is certainly on the first step in decay. Cracking should be observed. The more oil a varnish contains the larger the cracks. Fine cracks are of little account, because they do not prevent re- painting or revarnishing. Large cracks, however, are a serious matter. It follows that the most elastic varnishes, and those containing the most durable gum, have the disadvan- tage of producing, when they crack, the largest and longest partings of surface. This is the case with English varnish. No varnish is so durable. It is elastic, and retains its lustre for a long time — longer than any American varnish. In winter it dries to a fine, hard, very brilliant and durable surface; but in the warmer months, it dries with a " tach," on which one might gild. English varnish is therefore not fitted for summer use. Again, one must know what one puts over English varnish. It will do with its own, but if even two, or two and a half years after a coat of English varnish has been used, an American varnish be put over it, there is a large percentage of danger that both will crack in large cracks, running in several directions, but with a tendency to change to a direction at right angles with the grain of the wood. Whoever varnishes or repaints a car or carriage on which Eng- 26 Painting and Painters' Materials, lish varnish has been used should know that fact, and it would be advisable in buying a car or carriage to get the information at the time. An English varnish needs a strong and not too quick varnish over it, and is more dangerous as a basis for repainting than American varnish. Varnish-makers think and speak most of perishing; but this is not the worst of evils of varnish, because it is so easily cured by a second coat. A varnish should, however, hold its lustre for ten to twelve months. On the whole, therefore, while American varnishes "perish'* more quickly, more quickly lose their lustre, have less oil, are less elastic and less durable, they are to be preferred for general use. They crack in fine checks, neither large nor deep ; they are safer as a basis for repainting, and less dan- gerous to revarnish. CHAPTER IV. The Use of Varnish. A varnish may be called a liquid glass, and has much the same use as picture glass. A glass pre- serves the picture beneath it from water, dirt and injury ; a varnish does the same. Mulder, the able Dutch chemist, about whom we shall have much to say, recommends covering oil paintings with glass as the only effectual method of preserving them from the action of the oxygen of the air. Varnish over a picture preserves the oil colors beneath it, but is itself affected by the oxygen from which it protects them. For this reason it is much less durable than glass and may crack and injure the picture, or become cloudy and hide it. When a varnish over a picture becomes white and cloudy, it is generally because the varnish is no longer a continuous sheet like a trans- parent glass, but has changed into a fine powder, like powdered glass. A powder throws light in every direction, and is therefore not transparent ; a sheet of glass throws light in several regular ways, so that, by standing in a certain position and looking at a glass over a picture, we get the light reflected from its sur- face and see only a glassy lustre, through which we can barely discern a dim shadow of the picture. To this peculiar manner of reflecting light, glass and varnish owe their power of bringing out colors. When we get the glossy lustre from glass or varnish into our eyes, the picture beneath seems dim because there is so much of this white glossy light mixed with 2S Painting and Painters' Materials. its light. We are getting all the light there is from the picture, but our eyes do not see it because they are blinded by the brighter, glossy light from the sur- face of the glass. By taking another position, we see the picture very clearly, because our eyes no longer get the glossy light from the glass. Glass and varnish over colors bring them out (as we say in familiar language) by leaving them alone. Glass and varnish throw the white, glossy light out of the way, and the picture appears brighter to our eyes because it has less white light mixed with its light colors and shades than a picture over which there is neither glass nor varnish. This power of varnish, its ability to bring out colors beneath it by leaving them alone, unmixed with white light from the surface, is a very important fact. We place varnish over finely colored woods in The Effect of Varnish. 29 order to see more deeply and clearly into their struc- ture by getting rid of the strong surface light which blinds our eyes to the feeble rays coming from below the surface of the wood. The glitter of varnish we do not want ; it is hard, cold, meaningless light. The effects produced by the varnish are what we wish, be- cause it enables us to see deeper than we otherwise would be able to do. Varnish, therefore, adds depth to a surface. A well-varnished surface is the oppo- site of a flat, dead surface. This impression of depth is both a fact and an impression of our minds ; we the contrast between the surface of the varnish and the deeper surface beneath it, but the difference be- tween the two seems much greater than it actu- ally is. Therefore in using varnish there are two effects to obtain and one to avoid. 1. To bring out wood or paint in its own light by removing out of the way the white light which comes to our eyes from the top of the surface. Varnish by its lustre throws this white light out of the way for us ; the lights and shades of the wood and the colors of the paint seem brighter because they are purer. 2. To give a pleasant illusion of depth to a surface by the contrast between the upper surface, or varnish light, and the lower surface, or color light. This depth gives a softness to color which is the exact opposite and contrast to a glitter of varnish.* 3. There is no doubt that a gentle twinkle is part * There are other facts in connection with the relations of varnish to light which might be added, but they are more difficult to comprehend. Varnish as the medium of light has an effect. Von Bezold and Rood discuss the matter from different standpoints. JO Painting and Painters Materials. of the attractiveness of a varnished surface, but the effect to avoid is throwing the gHtter of the var- nish into the eyes of the observer. Great care is taken in this respect, in hanging pictures, to get a position for them which will throw the glitter of the varnish quite out of the way. The only care which the painter can take is to reduce the glitter to just power enough to accomplish the desirable effects (i and 2) without dazzling the eyes of the observer with the very white light, which it is his purpose to throw out of the way. Finish, therefore, must be adapted to the amount and kind of light which it is to get. A darker room or a darker surface will bear a higher finish than a lighter room or a whiter surface. The principle is that effects Nos. i and 2 are always desired, and No. 3 always as far as possible to be avoided, and in this matter the angle of light falling upon the surface has, of course, great influence. Rough- s tic ff and Rubbing Vajyiish, — Varnish has to do almost entirely with our eyes ; it is a matter of light and not of ^touch. It is for effects of light that so much trouble is taken with rubbing var- nish and rough-stuff. It is all a matter of throwing white light. If the white light is thrown in many different ways, from different spots, the surface ap- pears full of hills and hollows. If the white light is all thrown in certain directions, the surface appears smooth ; if^the white light is only thrown out of the way in spots, the surface is broken into blotches of glitter and dull color ; if the white light is thrown entirely out of the way, the deep, clear beauty of the surface beneath pleases the eye. The question arises whether too much trouble and The Effect of Varnish. 3^ too much expense and risk are not often involved to produce a temporary effect of this kind. Why, for example, should all this trouble be taken with the surface of cars, which are to be almost constantly covered with cinders and dirt ? It seems an unreasonable proceeding to spend weeks, or even days, in getting ready a surface which shall throw white light quite out of the way, and then as a light reflector use a cheap Kauri varnish, whose lustre dies in a few months, and with it, of course, those effects for which so much trouble has been taken. Still more unreasonable is it to endanger the protection and appearance of such an exposed surface as a car for the sake of a little more regularity in getting rid of a little more white light — for this is what is accomplished and about all that is accomplished by the use of rubbing varnish. Truly, it seems time that practical men should look at things as they actually are. Moreover, it is possible to give even a carriage so much glitter of finish that the eye is offended. Rubbing Varnish. — Varnish depends for its elas- ticity upon the amount of oil it contains. Rubbing varnish must be not merely dry, but hard, within 24 hours after its application. It, therefore, contains little oil, and this is made into brittle soaps by strong driers. No coat, therefore, has so little elasticity for its bulk. But this is not sufficient ; this non-elastic coat must be rubbed with water. Water evaporating upon a var- nished surface tends to separate the gum from the oil, and thereby to decrease its elasticity still more. See Apparent Changes in Figments due to Changes in Oil J2 Fainting aiid Painters' Materials. One of the difficulties in the way of getting rid of it is the danger of striping on flat color, but this may be avoided by using a coat of color and varnish, rubbing this with a bunch of hair, and striping upon it. vs Where rubbing varnish must be used, it is well (knowing each varnish and testing the mixture) to mix rubbing with finishing varnish, half and half, or less of the finishing than of the rubbing if preferred. A rubbing varnish should fill the following require- ments : (i.) It must be sweat" — /. it must remain flat and lustreless after the lustre has once been removed by rubbing. (2.) It must dry hard and firm within 24 hours, in order to cut freely and easily when rubbed with pumice stone and water. CHAPTER V. Testing the Qualities of Common Pigments. The usefulness of a pigment in relation to the oil will be thoroughly discussed; but we may first ex- amine those qualities which have not to do with pro- tecting, but with coloring properties. These qualities are (i) covering (coloring) power, (2) kind and quality of color, (3) durability of color, (4) price. Covering {Coloring) Power, — Cold-blooded people paint not to satisfy their sense of color but their sense of fitness. Color," says the first decorator of our time, " will never take real hold on the art of our civilization. Imitation and affectation may deceive people, but the deception will not last. To have a meaning and make others feel and understand it must ever be the aim and end of our Western art." Nevertheless, color is more important than any other quality of a paint. In America, until now, we have painted with white as expressive of the strong American sense of neatness, but the fashion threatens to turn bilious." Many white substances are white because they are in fine particles. A white lily is white because it con- sists of little cells which reflect all kinds of light, again and again, until it reaches our eyes from some part of its surface. Water becomes white when it is broken into fine drops, as in a waterfall, or on the crests of waves. White lead and zinc owe their white- J4 Painting and Painters' Materials, ness to their dense, fine, powder-like condition. Finally, transparent glass becomes white when it is ground into a powder. Covering power is due to two qualities, (i) White- wash made of lime and water has very little covering (coloring) power until it becomes dry. Barytes covers well as a water paint because, as in the case of white- wash, the water leaves it as a dry powder, but it covers poorly in oil, because the oil remains with it, and the light reaches it through the oil. Prof. Von Bezold has illustrated this curious and important question in the following manner* : Experiment^ illustrating how lime^ barytes and white leads containing crystals become in oil more or less trans- lucent^ and therefore do not color the surfaces on which they are placed. " If we fill the lower part of a small glass tube (a test tube) with coarsely powdered glass, the powder will appear white, and it will be impossible to see through it, but as soon as we pour water into the tube the powder will become translucent to a certain degree. By substituting turpentine for the water, the degree of translucency is considerably increased. Furthermore, if we add a small quantity of sulphuret of carbon to the turpentine, we shall obtain a liquid which reflects (bends) the light about as powerfully as glass, and if we now pour some of this liquid upon the powder, the latter will disappear almost entirely to the eye and we shall be able to look through the glass freely as if it contained only the clear fluid with- out the least particle of powder. If we immerse a glass rod in such a liquid (instead of which we may * Von Bezold, ''Theory of Color," L. Prang Co. Testing Pigments, also employ a mixture of olive oil and oil of cassia) it will appear as if the rod reached only to the surface of the liquid. Within the liquid the presence of the rod cannot be detected; it is perfectly transparent, as shown by the illustration. Instead of the powdered glass, small beads of transparent, colorless glass may be used. They will become visible as soon as the liquid dislodges the air between them. *'It is shown by these experiments that the presence of one transparent body within another is only be- trayed to the eye when the two differ in their power of refracting light. If this is not the case, the light passes through the mixture without obstruction." GENERAL TEST FOR WHITE PAINTS. (Chandler.) " Any white paint ground in oil may be tested by comparison with a pure article : " Weigh out loo grains of each paint to be com- jd Pamting and Painters Materials, pared, add 3 drops of linseed oil to each, and spread as nearly as possible alike on 6 x 12 glass,using a steel spatula for this purpose. Place the samples (thus prepared) between your- self and the light ; the sample which shuts off most light, and appears darkest, has the greatest covering capacity and its purity may be inferred as it excels in this respect." There are other questions connected with covering power which are too difficult to explain here ; but the above facts will aid in making it clearer why some pigments lose their covering power in oil. White lead and zinc owe their coloring power partly to the fact that they are dense metals which are in fine powder. White lead as it ages loses some of its carbonic acid and covers less well than at first, so that a black color under it is not so much hidden as before. This is also to some extent the case with zinc white, but it is evident from the fact that the hydrated oxide of lead (see White Lead) has no covering power, that it is to a change in the condition of the lead by the carbonic acid that the whiteness is due. This change may be merely the reducing the lead to fine particles whose surfaces are very irregular, and therefore reflect light like so many pieces of glass ground to a powder. Cer- tain it is that large particles of white lead whose sur- faces are more regular and crystalline* do not cover so well. (2) It seems certain, however, that the weight and smallness of the particles are the important facts. One of the best authorities says that, weight for weight, zinc covers as well as, if not better than lead but in a * From a Greek word for frost, reflecting light like ice. Qualities of Figments. 37 greater number of coats. The same is true of the different sorts of white lead. Mulder, the able Dutch chemist, estimates that three coats of lead are equal to five coats of zinc ; differ- ence in covering by these substances, therefore, is merely a question of labor ; with fewer coats one can get a better covering, but will use the same weight of metal. There is, however, another sense in which cover- ing power is used, which is entirely different from the above. Zinc white is said to cover 33 per cent, more surface than white lead ; and a good iron paint (as for instance the English Torbay paint) 62 lbs., as much surface as 112 lbs. of either white or red lead. Truly, but zinc white covers its one-third greater space with a thinner layer ; and likewise with iron paint. Thickness of layer as a protection to iron must be taken into account. Covering power has, therefore, three or four senses; protection to the oil, fullness of coloring to the sur- face, the amount of surface colored, and the thick- ness of covering resulting from the union between the oil and the pigment. The kind and quality of the color is the second and all-important quality of a pigment. It decides against the use of red lead, which in many respects is the most valuable pigment there is. It also decides against the iron paints, which are in other respects the best of paints. He is the skillful surface painter who can produce cheap, quickly-drying, hard, durable paint of clear, deep and quiet hues. In artistic painting, the problem is to produce last- ing and durable hues which shall be clear and pure, free from all muddiness and change. J 8 Fai?iting and Painters' Materials. Durability of Color. — Zinc white is the most dura- ble of white colors, sulphur gases changing it to a white sulphide of zinc, while the sulphide of lead is black. Zinc white has also less tendency to become yellow than white lead, although it darkens in the shade to a degree. Lead grows darker and more yel- low with each year, wherever exposed to sulphur gas or deprived of the sun. Lead tends to powder, zinc to flakes or scales. White lead, being carbonic acid lead, is not affected by carbonic acid gas ; the carbon- ic acid in rain water changes the zinc (oxide of zinc) into carbonic acid zinc. The acids of unseasoned wood also have a great effect upon it (Dent). Al- though the carbonic acid is easily driven off from the white lead by even a weak acid, yet it is doubtful whether zinc has more advantage than durability of color over lead ; and whether lead will not in exposed positions hold out as long or even longer than zinc. For the interior of the house, zinc is the superior paint over a foundation coat of lead. Zinc is imme- diately less bright than lead, but can be improved by a little varnish (dammar or other varnish) and in time becomes very hard and takes a good polish. Mixtures of zinc and lead are probably better than either alone, although one of the best painters in our acquaintance insists that it results in a tendency to crack. This opinion is not borne out by the experi- ence of others of less intimate but larger experience. It may, however, be true. Many zincs contain sul- phuric acid and, therefore, are in danger of darkening the lead (by producing sulphide of lead) or injuring its covering power by making sulphate of lead. Barytes {Sulphate of Baryta). — This substance is a Qualities of Pigments. jp natural product, known abroad as heavy spar and also as an artificial chemical product (permanent white). It is found in this country in Virginia and several other states, and is washed and ground and mixed with anything which will allow of the mixture. A " floated " barytes, /. a finer quality made by floating off and settling the finer particles, is also used. Artificial barytes has a greater covering power than the natural article, and is known as blanc fixe^ permanent white, etc., and with sulphide of zinc makes up the new Fulton and other whites, although the Charlton White Company has recently taken out a patent for the use of strontia instead of barium. Probably the larger number of tons of white lead used on this planet have contained barytes ; and as an honest and acknowledged adulteration not exceed- ing lo or 15 per cent., there is no proof that for out- side work it is not a gain to both durability and price. Zinc lacks weight, and this the barytes has, and it is a wise addition to zinc for outdoor use where lead is not preferred (Masury). Neither barytes nor any other substance which does not unite with oil is fit (by itself) for a paint. Lead does not cover so well with barytes, but zinc in the best sense covers better — /. e,^ protects better. I cannot speak from definite experience. As some pigments contain 90 per cent, of it, it is wise to listen to what can be said in its favor. One of the best au- thorities on paints, while he admits the injury to the covering power of pigments by an adulteration with barytes, gives it the credit of these advantages : (i) It brightens dark colors. (2) It injures chrome yel- low less than it does some other colors. (3) It pre- 40 Painting and Painters' Materials. vents pigments needing a large amount of oil to reduce them to the consistency of butter (as in pre- pared paints), from absorbing so much, by sooner producing this consistency. Its principal disadvantage is its lack of covering power, and the ease with which it induces men to in- jure their own character as well as that of their paints. The best houses, however, sell no adulterated white lead under their own names. An American chemist* who a few years ago inves- tigated the adulterations of paints in our market, gives the following as his results : A second-class zinc white : Barytes 54 Zinc white 44 Other substances 2 100 The adulteration of colored pigments by barytes is universal, but largely it is not dishonest, because the pigment is so made up for the price. There is, how- ever, dishonest adulteration, also, here. Iron Paints (oxides of iron mixed with clay, sand, etc., either naturally or with purpose). — No sub- stances are more lasting. There are houses in Sweden which, says the chemist Berzelius (1838), have stood well preserved for 300 years covered with iron paint. Mulder suggests that the pitch in the wood was not without influence in this case. Another au- thority suggests the use of fish oil either with or over the paint, or else the paint was a sulphuric acid iron, boiled with the oil, producing a preserving fluid which little by little sank into the wood. ♦Henry G. Debrunner. Qualities of Figments. 41 The only objection to these pigments for all out- side work is their color. Aniline colors have been used with them (as prepared paints), but are not to be trusted. Burning any suspicious paint over an alcohol lamp will destroy the aniline and leave the reddish iron in its natural color, exposing the cheat. Sulphuric acid iron colors, however, are well suited to this purpose on wood, less so for use on iron. These iron paints require a treatment which their price in this country hardly allows. They should be ground to the finest powder, and then ground again for some hours with oil of the best quality, as is the English Torbay. Unless they contain their full amount of oxygen in the natural state, they should previously be thoroughly roasted, when they become of a violet-red color. With age their tendency is to darken from loss of oxygen and the formation of black oxide of iron ; but their great durability is due to the fact that the air has so little effect upon their substance. Red lead is a valuable addition both as a drier and as supplying the qualities (except color) which iron paints lack. Iron paints are largely adulterated, naturally and otherwise, and the adulteration is an important con- sideration. Barytes, chalk and silica lessen the cover- ing power ; but clay has the disadvantage of being af- fected by water. The best adulteration is pure sili- con — quartz, sand, etc. Test is difficult except by analysis. A magnet may be used to discover whether there is any iron in the paint ; but the quantity of clay present is not easily ascertained. 42 Painting and Painters' Materials, Ochres. — These are clays naturally tinted with oxides and manganese containing water. They are the oldest and most lasting of pigments. Samples have been found in Pompeii ; they were known in Greece, possibly in old Egypt.* Their special value is in tinting, the best ochres, well ground and washed, having the brightest of durable tints, in yellow es- pecially ; /. e., speaking with a view to cost and dura- bility. Said Bouvier : " No other color can take the place of yellow ochre." There is, however, a great difference in quality of even French ochres, the lower grades containing more clay and being, perhaps, less desirable (on this account) than some American ochres. In general, ochres contain less than 40 per cent, of oxide of iron. A well-known yellow ochre found at St. Georges sur les Pres (France) is composed as follows : Clay 69.5 per cent Oxide of iron 23 5 ' Water 7.0 " " The burned ochres change to a variety of hues ; those containing more iron changing to a reddish brown ; those containing more manganese, to a chestnut brown. Therefore the darker ochres dry better. The change is due to a loss of the water and to the linking of oxygen with the metal (in place of the water), especially so in the case of the manga- nese. The ochres are permanent to a great and last- * The four pigments used by the Greeks in their pictures are said (Plinyj to have been white, yellow Athenian ochre, red ochre from Sinope, and black. This corresponds very closely with the palette used by Titian (chief of all colorists) for his dead color. He knew that his red was an ochre, and, therefore, the yellow was also an ochre for harmony. Qualities of Pigments. 43 ing degree ; but still the light blackens yellow ochres somewhat. The adulteration of ochres and of all mixed and more common pigments is an important factor. Lime injures their covering power, as does barytes, for the reasons given above. Clay gives the pigment a softer and less gritty touch ; but clays are affected by water, unless, as they do, they soak up quantities of oil. It will be necessary to speak of tests at another time. Finger Test for Fineness. — All the above pigments, however, may be tested as to their fineness between the fingers, or by rubbing them upon a stone with a knife. All gritty particles and large pieces indicate a badly washed pigment. Fineness of particles is a very important quality. Terra di Sienna, — The true article is a valuable pigment, brighter and more transparent than ochres ; raw, it has " a yellow-brown hue, producing with white bright, sunny tints burnt, it becomes a rich orange russet, more transparent on drying. It is an iron clay, and its superiority of color may be due to a small quantity of sulphuric acid in its composi- tion. The native American sienna is entirely in- ferior. The raw sienna requires 33 per cent, of oil to pre- pare for market ; the burnt sienna, 25 percent.; there- fore, it needs turpentine only, and will bear much thinning with it. So thinned it dries well. Umber. — Much said of the ochres will apply to umber, which is a manganese ochre from the island of Cyprus ; there is also an American article. The Cyprus umber is a soft brown color, " lovely raw umber" 44 Painting and Painters' Materials. (Samuel Palmer), ^* one of the most delicate of all the earths" (Hamerton), but with white, burnt umber gives muddy tints. The white seems to reveal in them (umber and Vandyke brown) possibilities of disagreeableness which were unsuspected when they were alone. Vandyke brown and white look like a mixture of chalk, mud and the lees of wine ; they bear no apparent relation to the fine, deep, semi- transparent brown color which bears the name of the illustrious artists." Burnt umber and Van Dyke brown, however, are useful house tints ; but with white lead theircolors become spotty by the powdering and wash- ing of the carbonic-acid lead. A little burnt umber with white lead grows lighter with age ; more gives a continually darkenifig tint, partly due to soap-making. By itself it has a good body, and stands perfectly. White Lead. — White lead owes its great useful- ness to its density and its relations to the oil. Weight for weight zinc white may cover as well as white lead ; but three coats of lead are equal to at least five of zinc. White lead should, for this and other reasons, be mixed thinner and zinc thicker than is the custom. First-grade lead from well-estab- lished houses is usually pure ; second-grade lead may contain one-half barytes, and third grade be com- posed entirely of zinc white, barytes, sulphate of lead, lime or chalk, etc. Adulterations. — Second-class white leads are fre- quently met with containing from lo to 50 per cent, of white lead, the remainder being zinc white, sul- phate of lead (which is white but has little cover- ing power), chalk, whitening, gypsum, barytes, clay, etc. Qualities of Pigments, 4^ A third-class white lead mdiy be represented by the following sample :* Lime 14 per cent. Zinc white. . . , 60 *' Barytes 20 ** Other substances in small quantity 100 percent. This "lead" was probably made by grinding together 25 parts of barytes, 15 parts of whitening and 60 parts of zinc. Lime is the most injurious and dishonest adultera- tion because it turns the dried oil paint yellow. Be- sides, it greatly lessens the covering power of lead, as do also sulphate of lead (white), barytes and zinc white. Zinc white contains, in its cheaper forms, sulphuric acid, which may blacken the lead and other- wise injure it. Test in oil is difficult ; in powder less so. Jn Powder. — Weigh a small quantity of lead and heat it red hot ; the carbonic acid is driven off with a loss in weight on an average of 14 grains in every hundred. As much as 16, or as little as 13 grains may be lost m weight ; but if very much more or very much less than these amounts the lead is proba- bly adulterated. Test No. I. — Any white paint ground in oil may be tested by comparison with a pure article. Weigh out 100 grains of each paint to be compared, add three drops of linseed oil to each and spread with a steel spatula (or knife) on sheets of glass 6 x 12, as nearly as possible in the same manner. Place the samples (thus prepared) between yourself and the light, and * Debrunner. 46 Painting and Painters Materials. looking through them, you will have no difficulty in deciding which is the most opaque, and therefore will cover best, and is (probably) the purest lead. The sample which appears darkest and shuts off most light has the greatest covering capacity. Cha7idle7''s Test. — Use 100 grains of white lead ground in oil, ^ grain lamp black (best), 4 drops of boiled oil. Mix thoroughly on a glass plate with a steel knife, and you will have the materials for a quantitative" test : Paint the mixture on a surface, the resulting color tests the purity of the lead. White lead With barytes mixed as above. in oil. Color produced. 100 grains and o grains Light drab. 95 " 5 ' Slightly darker drab. 90 *' 10 50 " 50 " 00 lamp black gr. 100 " Black. *' On trial, six different pers©ns agreed separately to the differ- ences in color as given above." (Prof. C. F. Chandler). If zinc white be in the lead, the color produced by the mixture will tend to bluish drab in proportion to the amount of zinc ; 6 or 7 per cent, of zinc is sufficient to give a decidedly bluish tint. On the other hand, mixture with barytes gave a pure drab, darker in proportion as adulterated with barytes. The practical painter will have no difficulty in apply- ing this test with sufficient accuracy, if he will weigh out in ordinary scales, say 100 ounces (6)^ lbs.) of each sample to be compared, adding to each half an ounce of dry lamp black, and to each sample an equal quantity of boiled linseed oil. After mixing the lead, black and oil together, very thoroughly^ spread each sample on glass, wood, or other smooth surface as Qualities of Pigments, 47 nearly alike as possible, when the difference in depth of color produced by the black will determine the comparative value (or covering power) of each sample. The sample most discolored will have the least body, and that least discolored the most body. (Prof. C. F. C.) White Lead. — The table on the following pages may assist those who desire to discover what adultera- tion has been used in lead or other white paint. 4^ Painting and Painters' Materials. Nameofthi Pigment. Conduct toward Heated be- fore the blow-pipe. Special proper- ties. Muriatic acid. Caustic soda. I. Whiting ( p r e c i p cnalk, Vi- enna chalk, Blanc d e T r o y e s, Blanc d e M endon), calcium car- bonate. Soluble with efferves- cence. 1 u n- changed. Becomes in candescent and turns turmeric paper brown after cooling. - Not poisonous. 2. White lead ( c e r u s e, pearl white, H a m b urg white, Ven- ice white, etc.), car- bonate of lead, etc. Soluble with effervescence and deposi- tion of small crystals. Soluble without residue (80 per ct. when bad.) A coating formed on the charcoal, citron-yellow when hot, sulphur-yel- low when cold ; easily fusible metal- lic beads also formed. Blackened by sul- phureted hydrogen. Poison- ous. 3. Pat ti son white lead, lead oxy- chloride. Soluble with- out effer- vescence. Deposition of small crys- tals. Same as above. Same as above. Same as above. 4. Zinc white, zinc oxide. Soluble. No effervescence. Soluble without residue. Yellow wliiit hot, white when cold. Fused with co- balt ni- trate solu- tion turns green. Somewhat Doisonous. Poison- ous. 5. Antimony white, anti- m n i ou s acid. Same. Same. White, easily volatile coat- ing, metallic globules which give off white smoke. White Lead Adulterations. 49 Name of thi^ Pigment. 1 Conduct toward Heated be- fore the blow-pipe. Special proper- ties. 1 Muriatic i acid. Caustic soda. 6. Bone ash, carbo n a t e and phos- phate of calcium. Soluble after heating effer- vescent at first. Un- changed. Unchanged, but becomes incandescent. Not poisonous. 7. Baryta white (blanc fixe, min- eral white), sulphate of barium. Unchanged. Same. After ignit- ing, if moist- ened with muriatic acid gives odor of sul- phureted hy- drogen. Very heavy. Not poisonous. 8. Gypsum (alabaster ), h y d r a t ed sulphate of calcium. Same. Same. Incandes- cent, other- wise, like heavy spar ; gives water if heated in a tube. Di^^icultly soluble in water. The solu- tion is ren- dered tur- bid by chloride of barium so- lution. Not poisonous. 9. Clay (china clay, etc.). Same. Same. Same. 1 Clay, moistened with cobalt solution and heated before Dlow-pipe, colored 3lue. 'J ale has a soapy feel- ing, is scaly in- structure, >feither is Doisonous. CHAPTER VI. Priming Wood. Hartig has estimated the room occupied by green wood to be as follows, per i,ooo parts : Fibre-stuff. Water. Air. Hard green wood 441 247 312 Soft " ' 279 317 404 335 395 330 300 370 A certain amount of water (7 or 8 per cent, of all) is included with the fibre-stuff. This shows us that about one-third only of the mass of the wood is solid stuff ; the remainder is either water or air space. When we dry out the water from the wood it shrinks (see tables on page 8), contracting about o.i per cent, in length and 5 per cent, in breadth, or 10 per cent, on the line of yearly rings. As we have already shown, this contraction is usually a matter of no great consequence to the painter except as pulling out into view edges of surface which may not have been painted. The important thing for the painter is to keep out the water after it has once been removed, because if the water gets in again it will (i) ex- pand the wood and crack the paint, with the grain ; (2) it will rot the wood under the paint. If the contraction of the wood is a matter of no great consequence to the painter, however, the pres- ence of moisture in its cells is a matter of the greatest consequence. JV/iy ? As is well known, wood is composed of cells like Priming Wood, those shown in the accompanying illustration. These cells are made of wood-stuff, which is not a solid mass, but itself composed of minute particles. Water sepa- rates these particles from each other and softens the stuff. Only in water-softened wood can the little plants which cause the dry rot get a foothold; they apparently feed on the minute particles of the fibre- stuff which have been softened in water.* Now the union between the particles of wood fibre and of the oil in paint is not a chemical one; there is no linking together, only sticking together of particles. Water does not to any degree affect paint, but it softens the particles of wood and the paint loses its grip on them. * Cellulose, of which wood, especially the outside of the fibres, is largely- composed, is much like starch, from which it is made and into which it can be changed. Cellulose, starch and sugar are all dissolved by water in increas- ing quantity, sugar being, in fact, the circulating form of cellulose. Fig. I. ^2 Painting and Painters' Materials, It is the same principle applied to the minute parti- cles of the wood which we work upon when we steam wood in order to bend it. Steaming wood will not, however, bend it, nor will mere wetting peel off paint; we must have a force. This force is steam generated in the wood by the heat of the sun or other heat rays. If any one desires to prove these statements, place an end of a painted stick in water, the end surface, of course, not being covered with paint. After a time make an attempt to peel off the paint at both this wet end and the dry end with a sharp knife. The result will show how much the wood immediately underthe paint has softened, and how much the paint has already lost its hold. A piece of spoke peeled in this way (the paint com- ing away almost of itself) shows by the ridges on its surface (the surface attached to the wood) what its attachments have been. One needs for the study a good microscope, with a variety of eye-pieces and lenses of considerable power. The lower powers will show long lines of ridges corresponding with the val- leys between two fibres. These are strongly marked; much less strongly marked are other very slight ridges crossing the surface of the paint, and corresponding to lines between, or on, the short medullary rays (c. c. hg. 2). With more powerful glasses, and looking deeper into the surface of the paint, molds of individ- ual wood cells (c. c. fig. i) are seen, and the imprint of their surface. All these lines and ridges tend in one direction (except those corresponding to the medullary rays), and this direction is z£////^ the grain of the wood. Any strain or movement of particles will. Priming Wood. jj therefore, tend to accumulate on this line, which will open cracks across the grain. The attachment to the paint is to the particles of the wood-stuff, as is shown by the adherence of certain portions after the paint has been soaked off. It is also held fast by being jammed into such valleys as we have described; but since oil contracts in drying, this hold is only occa- sionally a strong one. To protect the paint we must protect the wood par- ticles to which it is attached. We can best do this by filling the cells to which the wood holds ; a thin oil will soak their surface, but it is difficult to fill them. I have accomplished it with varnish, subjecting the surface to very strong heat for many days. The var- nish does not then soak loose because the water does 54 Painting and Painters Materials. not reach the particles on the wood surface, to which the surface of varnish holds. P7'iming. — The object of priming is to prevent the surface cells of the wood from being softened by water, or to protect the surface cells of wood immediately under the paint from water coming at them from below. For this reason a priming coat should be thin enough to enter into the ^Cox^-stuff, and surround those particles on the immediate surface. Priming is usually not a question of soaking wood cells, but of soaking wood-cell stuff. A primer must be a thin oil which will physically combine with the wood-cell stuff. It must not be made thin with turpentine, because turpentine evaporates, and we desire an oil which will stay, in order to protect the fibre from water ; and raw linseed oil unmixed with anything will serve. A little drier in hasty work is, however, allowable ; but it must give a thin oil. In case the wood is wet, we must heat the oil — not hotter than 150°. Linseed oil at 45° of temperature is not so fluid as water, but at 79° it is more fluid. The figures are as follows : Linseed oil. Degree of fluidity. Temperature. Water = 100.0 45° 86.5 79° 102.2 Or, to put the matter in another way, if linseed oil at 45° temperature requires 104 seconds to flow through a given aperture, it will flow through in 84 seconds if heated to 79^. Drying oil remains at that point in the surface where it drys ; but not-drying oil (as some primers) not improbably continues to sink into the wood until Priming Wood, 55 it finally disappears. Some painters who have used patent primers of some kinds assert that they result in a re-appearance of the grain of the wood after finish. Others make no such complaint — facts must decide in each case. Peeling off from Primer. — It is, however, not merely necessary that a primer be used which will penetrate the wood, but that the substance used shall be one which will readily unite with linseed oil, and to which linseed oil will stick. For this reason there is prob- ably no better primer than linseed oil, not mixed with anything whatever. CHAPTER VII. Oils. A Little Dictionary. Oxygen. — One-fifth part of the air ; the burning, rusting, change-producing element of the air. We breathe oxygen ; get fire and light by causing it to unite with wood and coal ; it rusts our iron (the iron in our blood), and we live by it ; it rusts our iron bridges, and we die by them ; we dry " our oils and burn our oils by oxygen. There is no substance of whose effects we know more by experience. Carbonic Acid. — The little and heavy gas of air. The product of decay, it produces decay in all living things which do not destroy it ; not immedi- ately, but by shutting out oxygen. Living plants change it into starch, starch becomes oil, oil changes again into carbonic acid and water. Carbonic acid destroys lifeless plants (trees and wood) rapidly. Iron rusts quickly in carbonic acid with oxygen. White lead is the rust of carbonic acid and lead. It is a fact of daily experience. It is only the names of these things with which the fainter is not familiar. Glycerine — To be had of your druggist. It is made from oils and fats. Glycerine Ether. — Glycerine as it exists in oils, united with an oil acid. Oil Acid. — An acid which when linked with glycerine ether is known as an oil. Oil. — An oil acid linked with glycerine ether, and from which glycerine and soap can be made. * Soap. — An oil acid linked with soda, potash, lead, zinc, iron or some such substance. Some soaps do not dissolve in water. Free Oil Acid. — An oil acid unlinked from glycerine ether, but not united with another substance. Flying Oil Acid. — An oil acid which becomes lighter than air (due to action of heat or light) and flies away as would a gas. Oils are useful in painting because water has little or no influence upon them. They are valuable to the artistic painter because there is less difference between the shades and colors of dry and wet paints mixed with oil than between dry and wet water colors. What is an oil ? ing it into a soap. Oils, 57 We discover the answer on mak- Suet fat. "Stearic acid. Glycerine ether Soap. ^ Glycerine. ^ Sodic oxide. Water. ^ Caustic soda. To those who have never studied chemical formulae the above may look perplexing, but it will soon be- come simple after a little attention to the unfamiliar words. Beginning at the bottom of the diagram we see that caustic soda is used to put with oil to form a soap ; and that caustic soda has two parts, which are linked together, namely, sodic oxide and water. In making this soap suet fat, or, as it is called, stearine, has been used. It also is composed of parts linked together, namely, stearic acid and glycerine ether. From the mixture of suet fat and caustic soda soap results, which, again, is composed of two parts linked, namely, stearic acid and sodic oxide. But another substance, glycerine, is also formed of the glycerine ether, which belonged to the suet fat, and of water, which was with the caustic soda. Oils — that is to say, those oils of which linseed oil is one — are composed of an oil or fat acid linked with water-free glycerine — and in making any kind of soap the glycerine is set free, leaving the acid by itself or linked with something else. ^8 Painting and Painters' Materials. Fish oil is largely Olein. Oleic acid. (jlycerine ether. Palm oil, which is used in making fine soaps, is Palmitin. r ^ \ Palmitic acid. Glycerine ether. Linseed oil is a mixture of Linolein. Linseed oil acid. Glycerine ether, and of olein and palmitin. In linseed oil there are 80 parts of linolein (linseed oil acid, glycerine ether), 20 parts palmitin (palmitic acid, glycerine ether) and olein, etc. (oleic acid, gly- cerine ether). Soap, — Clear ideas about soap will be very useful in studying oils and paints. Hard soap is made of oil acids and soda, soft soap of oil acids and potash. In making soap, the potash or soda is dissolved in water ; and this water remains in the soft soap, while from hard soaps it is drawn off, carrying with it the glycerine which was united with oil acid before the soda or the potash displaced it and left it free to unite with the water. The following is an analysis of sev- eral kinds of soaps. Some soaps contain also 2 or 3 per cent, of glycerine : TABLE NO. I. Oil acid. Potash. Soda. Water. ^Marbled soap (tallow) 81.25 1.77 8.55 8.43 Castile soap 76.50 9- 00 14 -So Palm oil (yellow) 65.20 9.80 19.90 Soft soap (common) 42.80 9. 11 48.00 (white) 45. 8.50 46.50 Soap. jrp It will be seen that the oil acid, the potash and soda and the water vary in several soaps. The soap is valuable in proportion to the amount of oil acid it contains, /. when the soap is to be used for toilet purposes fi-ee soda or potash and too much of either united with oil acids injure the skin.* In washing one's hands, soap removes the grease and greasy dirt which water alone will not dissolve, be- cause water has little influence upon oils. Part of the soda of the soap unites with the dirty oil acids, changing them also into somewhat of a soap, which is easily dissolved in water. The other part of the soda remains united with the oil acids of the toilet soaps, and forms a lather useful by inclosing and carrying away all particles of dirt, especially greasy dirt, which have or have not been changed by the action of the soda. The action of soap, therefore, is partly chemi- cal (the union of the soda with the grease) and partly mechanical (the removal of dirt by the lather). It is the same kind of action which removes the varnish from a carriage or a car when strong soap is used to wash it. It requires time for the oil acid to unite with the soda either in soap making or in hand washing. If a little of the soda unites with the oil acid, there is a weak soap ; if more, a strong soap ; in each case, however, the soda has taken the place of the gly- cerine ether which was united with the oil acid. The glycerine ether has been unlinked from the oil acid and set free. Glycerile unites with water and be- *The Pennsylvania Railroad had their attention called to 55oap by the de- struction of the varnish on a car by a washing with soap. Dr. Dudley, their chemist, looked into the matter, and all soap purchased is required to con- tain less than oi x per cent, of alkali carbonates. do Painting and Painters' Materials. comes glycerine, which mixes readily with the remain- ing water. Lime soaps and lead soaps do not dissolve in water, and by making a lead soap by uniting oil with litharge, it is easy to get the glycerine from the water, because the lead soap falls to the bottom of the vessel, and leaves the glycerine in the water, from which it may be obtained by evaporating the water. If pure gly- cerine be desired, some sulphur gas (sulphuretted hydrogen) may first be passed through the water and glycerine that any lead dissolved in the water may fall to the bottom. The sulphur in this gas unites with the lead and forms a heavy black powder — a sulphide of lead, the same substance which is found when white lead turns gray (not yellow) after it has been exposed for some time to air. The air of cities, especially, contains more or less sulphur gas. Oil Acids. — It is plain that oil acids may exist in three states : (i) linked with glycerine ether, thus forming oils; (2) linked with such substances as pot- ash, soda, lime, lead, iron, zinc, magnesia, or manga- nese, forming soaps; (3) unlinked and free — /. e.y chemically free — from glycerile or soap-making sub- stances, or, in other words, as free oil acids. In pro- portion as oil acids are free do they tend to form soaps when mixed with soap-making substances. Oils — /. oil acids linked with glycerine ether — may be mixed with soap-making substances without forming soaps. If two kinds of oils are mixed, one oil may more readily change into soap than the other, when to the mixture is added a soap-making substance. The oil acids may change otherwise than by becoming soaps. They may become rancid^ or they may dry. CHAPTER VIII. The Drying of Oils, and the Drying Oils. All oils tend to change in the air ; those less pure changing most rapidly. Butter which has not been properly freed from buttermilk is a good example. This change is, first, the unlinking of the oil acid and the glycerine, setting the oil acid free. Still, exposed to the air, the oil may give a sharp, unpleas- ant odor and taste due to further changes ; it has become rancid. Rancid butter may be sweetened, however, by washing it with some substance which will form a soap with these rancid acids and remove them. Lime water is sometimes used for this purpose ; and the clearing up of a fatty linseed oil by litharge is an action of the same nature. In this case a lead soap is formed by the acid and the litharge. The drying part of the linseed oil — /. linolein (linseed oil acid and glycerine ether) does not become rancid. The same is true of the drying portion of poppy oil and of other drying oils. The drying parts of these oils thicken and change into sticky, elastic, or into leather-like substances, all of which are water- proof, and upon all of which water has only a small influence. The Changes in Oils. — The chemical changes in oils which become rancid and in oils which dry " are of the same nature ; the results are different because the oils are different. 62 Fainting and Painters Materials. The science of chemistry is founded upon weight. If a substance changes and weighs heavier than be- fore the chemist is sure it has gained something. If a substance changes and weighs less, he is sure that it has lost something. He then proceeds to discover what it is that has been gained or lost, and in this his scales are again useful in weighing other substances with which this one has been in contact, in order to discover which of them has taken the lost something, or has given up the something gained. Linseed oil in drying gains from 8 per cent, to lo per cent, in weight. A French chemist exposed some drying oil to the air. It weighed about 155 grains; after eighteen months, about 166 grains. It had gained about 1 1 grains, or more than 6 per cent, of its weight. In speaking of the drying of linseed oil, therefore, we mean merely a change from a fluid to a solid condition. There is a loss of something, as we shall see, but there is a gain which not only makes up in weight for this loss, but usually leaves the same quantity of linseed oil 10 per cent, heavier than before it became "dry." The Influence of Darkness^ Light and Colored Light on Drying Oils. — A Dutch chemist, who has made a careful study of the drying of oils, arrived at the conclusion that linseed oil increases in weight in exact proportion to its becoming dry.* Without entering as yet into the question what it is that in- creases the weight of dried oil, we can study its dry- ing by comparing simply its weight. Any one may do this by comparing the following groups of figures with each other. We shall leave them in that measure of * Not absolutely but practically so, under the same conditions. The Drying of Oils, 6j weight which the chemist uses, namely, grammes. A gramme is 15.432 grains. TABLE NO. 2. Gained in Weight as Below. 3 d d d d G. 3 3 3 3 3 P a. a> CL fC CD a. rD Ou Oil drying AFTER. "kness . • uncolo: - red gl ' green * blue glass ' yellow red p glas ■ gla w Days. 0.000 0.126 o.oog 0.005 .023 0.089 0.012 20 .001 .258 .027 .245 .041 .002 .048 .076 •332 .103 40 .003 .326 .082 .139 •376 .184 60 .007 .298 .178 .269 ■ .388 •319 80 .013 .272 .284 .338 .354 •370 .388 .018 .261 .401 •357 .417 120 .024 •376 • 438 .360 .442 150 .035 ^300 •441 .485 .399 .474 Comparison of the increase by weight of the dry- ing oil under the several glasses gives the order of drying as follows : Most rapid under the uncolored glass, then follow blue, yellow, red and green, but this order is of no consequence, nor trustworthy as a constant fact*. After 150 days, on the contrary, the gain in weight is greatest in the following order, under green, yellow, red, blue and uncolored glass. In darkness there is no drying, or it proceeds so slowly that it may be said not to go on at all. Light is the first natural drier of oil. There is another experiment by a more accurate chemist, in which the gain and loss in the weight of the drying oil are given day by day. The oil was spread * See Changes in Pigments due to Sunlight. 64 Painting and Painters' Materials, upon surfaces, one exposed to the sunlight, the other remaining in half darkness. The measure of weight is in grammes and decimals of a gramme, as before. TABLE NO. 3. Oil Drying. In sunlight. In nalt darlcness. Date. Grammes. Orammes. April 2Q ..... ...6.860 April 30, g'd . . .03 Nothing. . .174 . .226 .004 4, . . .044 — 0.474 111 6 days. Nothing. — 0.008 " 9, " . . .034 .034 14, lost. . .041 Gained.. 1 80 20, " . • .030 " .378 28, g'd. . .026 " .220 June 4, .024 " .012 " 11, lost. . .043 Lost.. 028 " 18, " . . .014 *^ ..006 July 26, g'd. . .085 Gained . . 033 Aug. 19, " . 014 Lost. .003 Total gain. . 657 .865 Total loss. . . .128 •037 .529 gain. .828 gain. The powerful influence of light upon the drying oil is shown in the plainest figures by these two tables. They indicate that a painting or varnishing shop should have many windows, and a great abundance of uncolored light — no sort of driers in a shop are more valuable than windows. These tables also show that oil loses as well as gains something in drying, and that the least loss or greatest gain after many days is in slowly drying oil. Influence of Heat, — The following table shows the influence of heat upon drying oil : The applied oil was kept in a room having a tem- perature about equal to that of the Eastern States in the latter part of May and the early part of June The Drying of Oils. (from 59** to 68° F.). One of the oiled panels was subjected to a considerable heat for several hours each day (a heat of 176° F.).* Table No 4. Oil Drying. The panel not heated weighed 2.466 gramjnes. The heated panel weighed i-934 ist day — While heated for 2 hours gain .018 After heating " .072 .090 Panel not heated " .002 2d day — While heated for 4 hours gain .056 After heating " .065 .121 Panel not heated " .002 4th day — While heated for 6 hours 032 lost. After heating gain .013 Total 019 lost. Panel not heated . . gain .003 5th day — While heated for 7 hours 052 lost. After heating gain .017 Total 035 lost. Panel not heated gain .002 The panel which was heated but two hours the first day and four hours the second day — Gained .211 Panel not heated .004 In Other words, heating the panel dried it about fifty times faster than it would otherwise have dried, and that in a mild atmosphere. Every painting and var- nishing shop should be well furnished with steam-heat- * There are two kinds of thermometers, (i) Centigrade^ used by chem- ists and in Europe ; (2) Fahrenheit^ used in this country and in England. F." means the latter ; "C." means the former. The difference between the two is in the scale used to mark the degrees. 66 Painting and Painters' Materials, ing apparatus, or, if this is not possible, with a num- ber of stoves. As stoves are usually placed, one side of a painted or varnished car or carriage dries more rapidly than another ; there is need of heat on all painted sides. The heat, however, must not be too great, and es- pecially should it not be moist heat. A heat above 176° is probably injurious to good drying, but a less heat than this may injure the varnish or oil by driving it into the coats beneath or into the wood. This will be no injury to the wood (as we have seen), nor to the drying of the oil, but will tend to leave the " gums'* of the varnish on the surface. The temperature of a varnish room is usually 70° to 90°. The Gain in Weight of Drying Oil. — We have seen that oil in darkness hardly increases at all in weight. In the light, and especially if heated, drying oil be- comes from 8 per cent, to 10 per cent, heavier. Does it get this weight from the rays of light ? Science has shown that these rays have no weight, that both light and heat are merely waves of motion, like the waves which spread over the surface of smooth water when we throw a stone into it. The water is no heavier for the motion, it is only heavier because of the stone. Air, however, has weight as a stone has ; 100 cubic inches weigh 31 grains ; the air presses upon us on all sides, because it is heavy, so that a man's body sustains 37,560 pounds, or upward of 16 tons, of pressure from the air. Air, however, is not a simple substance ; it is com- posed of at least three gases, nitrogen, oxygen and carbonic acid. These are not linked together, but simply mixed Effect of Air in Drying Oil. 67 together as shot, peas and potatoes might be mixed ; so that while the names of these gases may be unfa- miUar, every one is constantly familiar with the gases themselves. We breathe into our lungs the nitrogen and oxygen (mixed). Air is nitrogen gas and oxygen gas mixed, 4 of N to I of O in every puff of pure air : N O N N N N N N N O Air from the lungs would contain all the ^' Ns," one less O " in every 5 " Os," and a little C, or carbonic acid, the heavy gas which causes clear lime water to turn white and milky on blowing one's breath into it. Every breath i7ito the lungs contains 80 parts of nitro- gen and 20 parts of oxygen. The air breathed from the lungs contains all the nitrogen (80 parts), but only 16 of the 20 parts of oxygen, and from 2 to 5 parts of carbonic acid. The other 4 parts of oxygen we keep. It is with the air as with some other things : We have had a large experience in it. We need only a little definite knowledge about it. Chevreul, a great French chemist, made the fol- lowing experiment in order to prove which of the gases of the air unite with the drying oil. He painted four panels alike, one side of each with white lead, the other side with zinc white, both prepared with lin- seed oil. No. I was placed in a closed glass box con- taining carbonic acid. No. 2 in a glass box contain- ing air. No. 3 was left exposed to the free air, and 68 Painting and Painters Materials, No. 4 was placed in contact with oxygen gas. The fol- lowing table gives the results : No. 1. Carbonic acid. No. 2. Limited air. No. 3. Free air. No. 4. Oxygen gas. JTl (Zi TABLE NO. 5. After 24 hours. The white lead nearly set, The zinc white The white lead nearly dry. inc white set but not dry. {White lead nearly dry. Zinc white set but not dry. The white lead perfectly dry Zinc white After 72 hours, but without adherence to the wood, absolutely fresh. Perfectly dry. Another French chemist analyzed oil before and af- ter its drying. He found ;* Before drying 11 parts oxygen gas. After 22 " Gain in drying 11 parts of oxygen. Otherwise there was little change in its elements, except a slight loss of carbon and hydrogen. As a rule, drying oil adds probably one-tenth to its weight by taking up oxygen from the atmosphere. It may add, however, even more, since there is no exact measure of the loss which is going on at the same time. What does Oil Lose in Drying? — Reference to tables No. I, 2 and 3 will show that oil loses in weight, as well as gains in weight, while drying, and that this loss is least in slowly drying oil. The comprehension of the process of this loss of weight will carry with it a better comprehension of the whole subject of drying oils. Black oxide of iron (black scale) and red oxide (red rust) seem quite unlike, yet they are both a union * The analysis, as we shall see, is not a close one. Losses of Oil in Drying. 69 of iron with oxygen gas. In the same manner, differ- ent substances result from the union of oil with the oxygen of the air. We have already seen that linseed oil is composed of Linolein. Linseed oil acid. Glycerine ether. Palmitin. A Palmitic acid. Glycerine ether. Olein. Oleic acid. Glycerine ether. Part of the loss in weight of the oil comes from the loss of the palmitin and the olein, the latter especially. When considerable oxygen unites with the oleic acid, a substance results which quickly flies away. When less oxygen unites with the olein, a more solid sub- stance results which does not so quickly fly away. Nevertheless, by heating the dried oil, we can drive it off from the still more solid dried linolein. As shown in table No. 3 (comparison of oil dried in sunlight and in darkness, page 63) oil dried in dark- ness gained the most. Gain in sunlight. Gain in darkness. ©•529 grammes. 0.S28 grammes. Heated to 176 degrees there were 0.106 grs. lost, .235 grs. lost, respectively, that in darkness losing at least one-half more than that oil dried in sunlight. Slowly-dried oil, therefore, contains more of the non-drying fats, and for this reason seems to gain (after a long time) more in weight. In table No. 2, (page 62), the quickly-drying oil weighed after 150 days 0.300, and the slowly-drying oil (under green glass) 0.485 grammes. YO Paiiiting and Painters' Materials, In the same way (by differences in drying) the dry- ing portion of the oil (the linolein) produces different substances, as we shall see. The Glycerine Ether. — There is also another and still more important loss of weight in drying oil. If we are so fortunate as to go to bed by the light of a tallow candle, we will observe a peculiar and sicken- ing odor on putting out the flame and leaving the wick to smolder. A tallow* candle is largely Stearine. Stearic acid. Glycerine ether. By the heat, some of the glycerine is unlinked from the fat acid and changed into the unpleasant, sickening substance which our nose discovers to be in the air. It is the same with the changes in both the not- drying and the drying parts of oil. The glycerine ether is unlinked from the fat acid and flies away. A heat of 176^ — much less heat than will boil water — begins to drive off the glycerine ether from the dry- ing oil. The process of drying appears (i) to begin with the unloosing (not unlinking) of the glycerine from the oil acid, thus : Oxygen gas. I Linseed oil acid. Glycerine ether. * Composite candles are made without glycerine, and give no such odor. The Drying of OiL 71 And at this point the oxygen of the air begins to unite with the oil acid to form a varnish.* The whole process of drying appears to be the linking together of oxygen with the oil acid, and therefore plenty of pure air is absolutely necessary to dry paint and varnish. The ventilation of a painting shop in order to get as much pure, warm air as possible without dust is a matter of the first importance. Abundance of light is of nearly equal importance, as we have seen. Be- fore drying, the oil is a fluid, water-like acid, the oxygen an invisible gas filling a large amount of space. After the drying, both appear together as a thin, solid, elastic varnish. Review of the Drying of Oil. — Although our knowledge of the drying of oil is still not complete, we may profitably take a review of what we have already learned. (1) . Linseed oil is composed of Drying oil 80 parts Not-drying oil 20 100 of which 8 parts are glycerine ether. (2) . Pure oil will not dry in darkness, light is neces- sary ; and the greater the light, the more vigorous the drying. (3) . Heat is also a powerful drier of oils. Oil heated for a short time continues to dry more rapidly for hours after the heating. (4) . Drying*' appears to be the loss of glycerine * This is an explanation by facts — for theory, see Theory of Linseed Oil, Part IT. ^2 Fainting and Painters' Materials, ether and the gain of oxygen from the air. The glycerine appears to be unloosed from the oil acid, as the oxygen unites with the oil. The result of the drying is a bright, hard and elastic varnish. (5) . Much of the glycerine ether which is linked with the oil acid flies away in the drying. (6) . Some of the not-drying oil acid also flies away in drying — especially under the influence of the direct sunlight, or when the oil is heated. less) THE DRYING OILS Lin(flax)seed oil freezes at i8.4°F.below zero Poppy seed oil freezes at 1-5' a u Walnut oil freezes at - - 18.4° a Prunella (a Japanese oil). Hempseed oil freezes at 18.4° a u Castor oil freezes at 1° a Sunflower seed oil freezes at - - 3.6° above n Grape seed oil freezes at 1° u a Cotton seed oil freezes at - 28° a i( Fish oils freeze at - - - 30° a u Oils are made solid by cold, and become less in volume ; in other words, they contract and solidify, and this is what is meant by the "freezing." One thousand and sixteen quarts of olive oil in summer (about 70°) become in winter only one thousand quarts (at 32*^). Those oils becoming solid at highest temperatures are most unfit for use in paint ; but results differ as to exact points, and the figures given in our table are all low points — some samples would become solid at higher temperatures. Some of the above are more properly half-drying oils. Linseed oil stands first in every good quality, except for its tendency to change color. Linseed Oil Of the family of flaxworts, there are several of this country; but the common flax plant* belongs to those old countries where men first lived ; and in all histori- cal times has been a cultivated plant. The flax plant, like all others, is a chemical labora- tory, the working-rooms being specially the leaves. In the seed are stored up the choicest results of the chem- ical work : (i) plant flesh (albumen), which exists in the seed as the body of a new plant : (2) sugar, starch, ucilage, which are ready preserved foods for the young plant ; (3) oil, which is concentrated or stored- up food, which may be (strange as it seems) changed into sugar or starch again before it is used. This oil is what we wish in painting ; the flesh (albumen), and the starch and mucilage are what we do not want. In pressing the oil, however, we probably get all three. A nalysis of the Ripening of Olives. Composition of pulp and fruit. , Percentages at different dates. > June 30. July 30. Aug. 30. Sept 30. Oct. 30. Nov, 15. Oil and green leaf substance. .40 5.49 29.19 62.30. 67.21 68.57 Some other sub- stances 98.00 94.00 52.00 27.00 22.00 •|'22.oo Percentage o f water in total plant 22.00 60.00 61.00 56.00 51.00 50.00 As we have no analysis of the ripening of flaxseed we give that of another oil-giving seed (the olive) at various periods of its growth. It will be seen at once that we should get more water and more substances which we do not want, and * Linum visitatissimuju. t These are woody substances, nitrogenous substances make up the re- mainder. The percentage of water is to the total plant ; of the substances above the line to the total of dr / plant-stuffs. I 14 Painti7ig and Painters Materials, Flax Plant — Flower, Seed Vessel and Seed. Its flower is blue. Fig. i represents a flower leaf or petal ; there are five to each flower, which is of a very regular and perfect kind, having five petals, five pistils, five stamens, five sepals. Figs. 2 and 3 are sepals, or cup leaves, to the flower ; figs. 4 and 5 represent the seed vessel, with its tall stamens and taller pistils ; fig. 6 is a stamen ; fig. 7 is a seed vessel cut open, showing 10 seeds. The stamens fertilize the pistils, the pollen falling upon the top of the pistil, or probably carried there by some busy bee. Within each of the pistils (not to speak exactly) grow two seeds, as seen in fig. 7, divided by a little wall. Fig. 8 is a ripe seed vessel. Sections of the seed and the perfect seed are seen in figs. 9, 10, 11 and 12. I Linseed Oil. 75 also much less oil, from green seed. It is doubtful whether either the plant-flesh or the mucilage would flow out when the seed is pressed, if it were not for the contained water. The Oil. — The process by which the oil is made in the plant is possibly as important as it is interesting to know. After years of study it is a matter of great uncertainty, but even out of uncertainties practical light may come. There is reason to believe that under the influence of the sun's rays the oil is made as follows : Under CD I Carbonic acid (of the air) becomes Green leaf substance, which becomes Starch, yellow leaf -stuff, which becomes Sugar, not drying oil (palmitin), which becomes Oil, drying oil (linolein) Everybody knows that grass is green, and that leaves are full of green stuff. Everybody knows that quantities of glucose are made from the starch of Indian corn. These bits of every day knowledge are sufficient for some compre- hension of how the oil is made in the laboratory of the plant. One of the most celebrated botanists does not be- lieve that green-leaf stuff is made from carbonic acid. 76 Fainting and Painters Materials, It is sufficient, however, if imperfect leaf-stuff be mixed •with linseed oil. The Mucilage. — We are soon acquainted with this pleasant slimy substance on putting some flaxseeds into the mouth, and perhaps we have also pleasant rec- ollections of soothing flaxseed poultices. It has the peculiar property of soaking up water, and is there- fore not desirable in the oil. The plant-flesh— the juices of the little crushed seed plant (see fig. 12 on page 74) — has the same influence on the oil which un- removed buttermilk has upon butter — namely, to make it rancid. It is doubtful how much of these substances are in the oil ; the chances are even that the foots " of well-settled oils consist of imperfect, unripe oil, mixed with the green-yellow plant-stuff from which it has been made.* I am told that more of these substances exist in American seed oils than in oils from seeds imported. The Calcutta seed oil is given the preference for color, both in America and in England ; and, although very little is imported, some of the white-lead manufactur- ers use it in grinding with the lead. The Calcutta seed, however, has probably no ad- vantage in the amount of mucilage it contains ; exam- ination by an English chemist showed quite as much mucilage and plant-flesh in these seeds as in others, f The difference in oils is probably in the amount of unripe oil which they contain, which is made from and * Mulder found no satisfactory evidence of plant-flesh in linseed oil, al- though he did not deny it might be present. The stuff thrown down from the oil by sulphuric acid and acetic acid lead was apparently heavy dark oil. + Mucilage in seeds ranges from i3 t ) 5.:> per cent.; albumen, from 17 to 25 per cent. Drying Oils. 77 mixed with color-stuffs. The amount of these foots in various oils has been estimated as follows : linseed oil i i-io parts in loo of oil. Poppy-seed oil and cotton-seed oil a very little more ; but all such estimates are difficult and uncer- tain as to what is found. The figures are valuable as showing how trouble- some may be a very small amount of imperfect matter. The Yield of Oil. — Heavy seed will yield most oil : and seed ripened under a hot sun, and when flax is not gathered too green, is the best. The weight of lin- seed varies from 48 to 52 lbs. to the imperial bushel, probably a fair average is 49 lbs., oil 7^ lbs. to the gallon — perhaps 3.60 lbs. of seed to one pound of oil.'* — (Woolsey in Ure's Dictionary.) OTHER DRYING OILS. Prunella^ a Japanese oil, has recently received some attention; but little, however, is known of it. Poppy Oil is obtained in quite large quantities from the small black seeds of the ordinary poppy plant (^papaver somniferuni)^ -which, are so rich in oil that they are capable of yielding about half their weight of it. It is clear yellow in color, but becomes solid at a higher temperature than linseed oil. In var- nish manufacture it is used for producing only es- pecially fine products, and it is also used by artists for diluting colors. Poppy oil dries faster than nut oil and slower than linseed oil. It, however, does not remain sticky so long as linseed oil. Both poppy oil and nut oil take up less oxygen in drying than linseed oil. Mulder speaks of the use of poppy oil by artists to 7

ench and English school of the last hundred years are of a still more seri- ous nature, and much more difficult to cure. Many of these, though never exposed to any injury whatever, cannot be guarded from pre- mature decay, in spite of all possible care with which they are kept. The principal symptoms of their bad constitutions are : (1) Darkening of the opaque-bright colors. (2) Fading of the transparent brilliant colors. (3) Darkening, and above all cracking, of the transparent dark colors. The best opportunity to study these several appearances is given us in the Museum of the Louvre, which contains a great number of such pictures in the section occupied by the French school. I have paid much attention to the cracks in these pictures, as I find that in shape, in size, in position, as well as in relation to the io6 Painting and Painters' Materials, various colors, they differ distinctly from the cracks in the older pictures and in those of other schools. The special characteristics of these cracks are the following ; They are all but exclusively found in the thickly laid on transparent dark colors, and they are the deeper and more gaping in propor- tion to the thickness of the layer of the color, and the extent of the dark surface. The chief cracks run parallel to the outlines of sur- faces painted with bright opaque colors, such, for instance, as are used for flesh tints, and which are more or less thickly laid on. But there is generally a slight distance between the bright colors and the cracks. Lateral branches of these cracks pass into white surfaces but they do not gape, provided the white colors had been laid on directly upon the priming and not upon a layer of dark transparent, but not sufficiently dry color. This examination of the cracks of pictures has sometimes afforded me a peculiar insight into the practice used for the pic- ture. In the well-known picture by Guerticault of " The Wreck of the Medusa," in the Gallery of the Louvre, the cracks follow exactly the outlines of the bright flesh tints. The arm of one of the dead bodies hanging in the water is so covered by planks and water that nothing of the forearm is seen. It is, however, very easy to prove that originally this arm was painted in its length, for the cracks do not only follow the outlines of the arm as now seen, but also the outline of the no longer visible forearm and all the five fingers. This proves that the fore part of the arm and the hand were originally painted in flesh tints before they were covered over by the paint of the planks, and the water painted afterward. In Ingres' portrait of Cherubini, the face of the latter is beauti- fully preserved, while that of the Muse, as well as her drapery, is covered with cracks. In the depth of the cracks of the white drapery an intense blue tint is seen. Ingres painted the head of Cherubini in Paris, and then took it to Rome. There it was pieced into new canvas and lined. Then the Muse was painted, and be- fore the colors were perfectly dry, another model was chosen, and a new Muse painted over the old one. The color of the drapery was likewise altered, and this explains the cracks in the white color, and explains also why the blue ap- pears in the depth of the cracks of the drapery. Among English artists of the last hundred years, some have painted wdth the same materials, and by the same process as the French, and with the same unfortunate results. Others avoided these by using the same materials with more precautions. The study of alterations already fully developed in the pictures painted within the last hundred years only, and their comparison with the works of the old masters, would suggest the following rules for the process of painting : ist. The oil should in all colors be reduced to a minimum, and The Cracking of Paintings. 107 under no form should more of it be introduced into a picture than absolutely necessary. 2d. All transparent colors, which dry very slowly, should be ground, not with oil at all, but with a resinous vehicle. 3d. No color should be put on any part of a picture which is not yet perfectly dry ; and above all never a quick drying color upon a slowly drying one which is not yet perfectly dry. 4th. White and other quick drying opaque colors may be put on thickly. On the contrary, transparent and slowly drying colors should always be put on in thin layers. If the effect of a thick layer of these latter is required, it must be produced by laying one thin layer over another, taking care to have one completely dry before the next is laid on. If transparent colors are mixed with sufficient amount of white lead, they may be treated like opaque ones. The varnish may crack or become dim ; it should then be treated by Pettenkofer's method;* but it may become dark yellow, brown and dirty, and so hide the picture as to require its removal, when it should be replaced by a thin layer of fresh varnish. If a picture is throughout painted in oil, if its substance has remained sound and even, and it was varnished with an easily soluble mastic or dammar varnish, then there will be neither difficulty nor danger in removing the varnish. The diffi- culty and danger are much greater in cleaning those pictures which have not been varnished with the ordinary easily dissolved mastic or dammar varnish, but have been painted over with oil, oil varnish (or oleo-resinous varnish). It seems incredible that these substances should ever be used for such purposes. It is, however, a fact that there are still people who fancy it will contribute to the pres- ervation of their pictures to brush, from time to time, a little of those liquids over their surface. They recognize too late that the varnish becomes more and more dark, of a brownish color and opaque. If such varnish has afterwards to be removed, then we meet with the great difficulty, that this can be done only with sub- stances which would just as easily dissolve the whole picture as the hardened layers spread over it. — Synopsis of an article by Dr, Liebreich. * See Appendix. CHAPTER XII. Manufacture and Use of White Lead. White lead, however made, is the rust of lead by carbonic acid gas, which may result naturally, but can be very much hastened by artificial heat and the artificial production of a large amount of the gas. . White lead also contains an oxide of lead combined with water. The Lead. — Silver in the lead produces a white lead of pinkish cast ; bismuth in lead gives a whiter lead, but may also darken it by collecting in quantity in some portions of the lead during the melting. The melted lead is, for the Dutch method, cast into the form of large buckles ; for other methods the lead may be used in the form of sheets or of powder. In this country, with the exception of three or four factories, none of which are large, lead is made by the Dutch method or process. Other methods are the French^ Ger?nan^ and modified forms of these. The Dutch method ^ so called because it was first used in Holland of all European countries : The buckles of lead are placed over pots, which contain vinegar acid (acetic acid of from .05 to .75 per cent, strength) the pots are placed in stacks" (or brick chambers) on a bed of hard, fermenting tan. The pots are quite close together over the whole bed except for a space of about 8 inches around the walls. This space is filled in with tan-bark, a board separating it from the pots. The first bed is then floored over, and a second bed made in the same The Manufacture of White Lead, log fashion on this floor. Ten or twelve beds are thus sometimes stacked and the top covered with a layer of tan bark. Some corroders of white lead use manure only ; others a mixture of tan and manure, both serving the same purpose, namely, to heat the acid in the pots. The manure, however, has the disadvantage of giving off sulphur-ammonia gas, which combines with the lead and produces black sulphide of lead. This is specially the case when the manure of pig- styes, and of flesh-eating animals, is used. The tem- perature of the pots reaches 140° to 200°. The quantity of lead in a stack of 2ox 12 feet is three tons, which lies upon 1,000 pots (each 5^ inches diameter) containing 200 gallons of vinegar acid. The vinegar acid is gradually changed by the air which, in small quantities, enters the stack, into car- bonic acid gas. At the time of stacking, the air will contain 20 parts in 100 of oxygen ; after a fortnight only 17 parts, and in five or six weeks 15, 13 or even but 7 parts of oxygen, the carbonic acid in the air having increased from 5-10 of i per cent, up to per- haps 27 per cent. A change in the lead before it is acted upon by the carbonic acid is required ; and this is, apparently, the reason why the process cannot be safely hastened by at once increasing the amount of carbonic acid in the air. The result of this impure air acting upon the lead buckles is white lead. Not all the lead, however, is changed to white lead, from 30 to 50 parts of un- changed lead remain, and must be separated. Of the white lead, 125 to 129 parts by weight represent 100 no Paintmg and Painters' Materials. parts of lead. When the stacks have been up from eight to nine weeks they are unloaded. The unloading, but more especially the separation of the white lead on the buckles from the remaining unchanged lead, is one of the most dangerous opera- tions in white lead making. In some countries this is done under water by means of rakes ; in others the dry powder is rolled and beaten on a stone slab with a sort of mallet, in which case the workman should have his mouth covered by a damp cloth or sponge to keep out the fine, dry white lead= After this separation process, the white lead is ground in mills, commonly with water, and either dried or in its wet state re-ground with oil. In this country some of the most careful manufacturers use oil from East Indian flaxseed. In other countries white lead is often sold in loaves made by binding the lead flour with gum arable, acetic acid, lead, etc. In this country it is usually sold in kegs, ground in oil, which is in every way an advantage to both the durability of the painter and the paint. It probably lengthens the life of the whole trade ; for grinding paint is one of the most common causes of lead diseases, especially colic, as shown by hospital reports and experience. The German, Austrian, or chamber process of white lead manufacture is an improvement upon the Dutch or pot method. The lead is used in sheets about I inch thick, 8 inches broad and 12 inches long ; 1,800 to 2,000 such sheets per box ; 8 boxes in each chamber, containing from 22,000 to 53,000 pounds of lead. The walls of the chambers are lined with tin, and the chambers heated sometimes The Manufacture of White Lead. iii with steam. The carbonic acid which rusts the lead into white lead is made by the fermentation of vinegar, yeast and other substances, ammonia and phosphate of magnesia being added to hasten the fermentation. Carbonic acid from burning charcoal is sometimes in- troduced into such chambers. The Kremser white or Klajigenfurt method is a German process, in which vinegar from dried grapes is used. The best of these leads are said to be whiter than Dutch process lead, and to cover equally as well. Flake white is a pure white lead in scales. Krems' or Crems' white is a poor sort of Kremser white. The Clinchy or French process is entirely different from the others. The white lead is made by passing carbonic acid (made from coke and chalk) through a solution of sugar of lead (/. some kind of acetic acid lead). The passage of the gas through the liquid lasts from 12 to 14 hours. The resulting white or carbonic acid lead (which settles to the bottom) is more or less crystalline, loose and coarse in grain. It allows more light to pass through it, takes up more oil, and does not, therefore, cover nearly so well as Dutch white lead. There are many different processes which may be classed under one of the above methods. H. Grun- eberg's, in Germany, and Milner's, in England, are worthy of mention as producing (it is said) white lead equal or superior (in whiteness) to Dutch process lead, without defects and with advantage of safety for the workmen. Milner's process is used by the Sankey White Lead Company, near Warrington, England. The lead may be made in two days, by the action of TI2 Fainting and Painters Materials. carbonic acid on a lead (oxy chloride) made by grinding together litharge, water and common salt. Gruneberg s process requires eight days, and results from the action of carbonic acid on very finely pow- dered lead and litharge, rolled constantly in cylinders. Pattisons white lead is an oxy chloride of lead, made by the action of muriatic acid on lead ore (galena). A practical question, as we shall see, is whether the difficulty with much white lead is not in pushing the process through in a short time. If, as seems probable, white lead is more apt to powder than form- erly, we are inclined to believe that here lies the fault. There may also be failure to wash the lead properly. The fresh lead contains loose carbonic acid and more or less acetic acid mixed through it ; these acids must be washed out. It is probably the carbonic acid which causes all white lead to powder. THE NEEDS OF A PAINT. A paint needs hardness in order to withstand the friction (i) of rain and storm, (2) of fine particles of sand or cinders blown against it by the wind, (3) and injuries and accidents of all kinds which tend to rub off and deface its coats. The oil itself is also better for protection from the oxygen of the air, which, hav- ing hardened it, tends to harden it too much, and finally to change it into carbonic acid and water.* An oil varnish may be called a hard paint. It is made so by removing the soft, non-drying oil acids by boiling the oil, and also by changing them into soaps. Finally, * As an illustration of the protection of mere hardness a freshly painted and varnished car will often suffer severe injury by immediate exposure to a hot summer's sun. Its paint soon thereafter perishes. Qualities of Durable Faint. iij the oil is hardened by mixing it with ^'gums." An oil varnish may be called a'' gum " paint, and the harden- ing of oil is always produced in the same manner, namely, by mixing it with fine particles of hard sub- stances. It remains to study the effect upon the oil of each kind of these substances — lead, zinc, iron, etc. There are those who are inclined to believe that all soap-making between pigments and oil is injurious to the durability of paint. There is ground for this in the fact that white lead, which forms some soap with oil, is less durable then iron paints, which are, to a greater extent, merely powders held in a coat of dried oil. It remains to discover whether longer life is given to oil by changing it into a soap. It is, how- ever, an advantage to get a hard paint immediately, if by so doing w^e can also retain the elasticity of the oil. This we accomplish by using red lead as a pigment. The red lead gives up some of its oxygen to the oil and changes part of it into an oxy-linseed-oil acid, with, which the lead unites to form a hard oxy-linseed- oil acid, oxide-of-lead soap. And yet red lead is a lasting paint, long retaining its elasticity and a certain power of holding itself together, more lasting than iron paints (see Painting Iron). We do not get this soap in the case of other sub- stances united with the oil, because neither the zinc, iron, and only part of white lead, gives up its oxygen to the oil, although they all contain oxygen, while the red lead also gives a hard substance with which to unite with the oil acid — /. e., it makes hard oxy-linseed- oil-acid, oxide-of-lead soap. Of these facts we may be sure, if we can trust so able a chemist as Mulder. 114 Painting and Painters Materials. Mulder s Experiments. — Having clearly seen the excellence of a red-lead paint, Mulder conceived the idea of making a good, cheap paint by combining (mixing) hard powders of cheap substances with it. For the oil, he took a boiled linseed oil, containing 2% per cent, of lead oxide; 100 parts of this oil to each mixture. The iron paint spoken of consisted of sub- stances as follows :* Water 2.75 parts / Red oxide of iron 68.27 ^ S ^''"^ ^^^^ ^* Clay 27.60 1 Marl, 27 1-28.27 parts clay. Chalk 40 J This was a purchased paint from a maker of iron paints, Cartiers, of Belgium. The other substances used were as finely powdered as possible. Two hun- dred parts of fine sand with 5 parts of red lead and oil gave a paint of some worth. Red lead 25 parts with 40 parts of iron paint gave a very good covering ; 20, 40, 65 and 90 parts to 100 parts of iron paint were ex- cellent. Twenty to 90 parts of red lead and 50 parts of red roof tiles gave a thick, heavy covering ; 40 parts of red lead and 100 parts of roof tiles gave an excel- lent covering. Finally, 20 to 90 parts of red lead to 1 00 parts of powdered ironstone gave a paint of dis- tinguished excellence. The iron plate on which these substances were painted was partly washed with sul- phuric acid before painting. After painting it was ex- posed to wind and rain until the unpainted side was covered with rust, but no trace of rust appeared on the painted side. After 10 months Mulder speaks of the undoubted success of the experiment. The paints were holding well and evenly as very hard, thick ♦ In the absence of American analyses of iron paints, the above is given in full as an illustration of their composition. Qualities of Durable Faint, ii^ coverings. Mulder thus sums up the method of mak- ing a cheap, hard and well-protecting paint ; 1. By boiling oil with 2 to 3 per cent, of oxide of lead (litharge or red lead) to harden the soft non-dry- ing and free oil acids. 2. Some oxide of lead as pigment : red lead the best (uniting with and giving additional hardness to the oil). 3. Hard indifferent powders, made as fine as possi- ble, and as much used as the mixture will bear as a good paint. The Theory of a Good Paint. — The above seems to represent the true theory of a good paint. We must have for a good paint substance one which not merely mixes with, but more or less links with, the oil. Barytes does not unite with the oil ; zinc does so slowly ; iron paints only partially, if at all — red lead is the substance which best fulfills this office, and it is the most lasting paint. White lead, as we shall see, owes all its value as a paint (not a color) to a substance like red lead, which it contains. Nevertheless, red lead is the most dangerous paint for the painter. Moreover, it is a difficult paint to obtain pure, and is more inclined to blister than are iron paints, which are also much cheaper. We do not advocate the use of red lead, but it is necessary to show the facts as far as they exist. The secret of a paint is to have some substance mixed with it, which forms a soap and hardens the oil. The valuable discovery would be a substance to replace lead ; until this is discovered, we advocate iron and zinc paints as far as they can be used, mixed with lead. Ii6 Painting and Painters' Materials. It is doubtful whether we can add lasting elas- ticity to a raw oil by boiling it, but we can possibly keep its elasticity longer and also harden the oil ; and get a quicker drying oil. Too much drier must not be used or the result will be too hard and brittle an oil, without elasticity. WHAT IS WHITE LEAD ? The soap-making of white lead with oil is certain, but not easy to understand. The following appear to be facts, based upon the experiments of Mulder and on those of two English chemists : White lead is composed of two kinds of lead. The whiteness and covering power are due to one, the strength and hardness of the paint to the other. White lead consists of carbonic acid lead and water lead. No. I, in table 14 below, is a good, dense lead, spe- cific gravity 6.32 ; No. 2, a dry white lead, specific gravity 6.50 ; No. 3 is a crystalline, transparent lead. TABLE NO. 14. No. I. No. 2. No. 3. 86.18 83.53 10.44 15 70 3-37 0.75 TABLE NO. 15. Carbonic Lead. acid. Water. White lead, best quality 86.80 II. 16 2.00 2d quality.. 11.68 1. 81 3d quality. . 86.03 12.28 1.68 14.10 0-93 83.47 16.15 0.25 White lead should contain : Oxide of lead linked with water, at least 25 parts. Carbonic acid lead 75 White Lead. 117 The carbonic acid lead gives the whiteness; the water-oxide of lead only hardness^ but without it there is no painty only a wash. The difficulty with a very hard white lead which con- tains no water lead is that it will not harden, but eas- ily brushes off like a lime whitewash. If abundance of oil has been used and this difficulty occurs, the painter may be quite sure that the lead has not been properly made. Carbonic acid lead produces some change in the oil ; it makes it into a sort of emulsion. To make the matter very clear, we will suppose that this emulsion, like other emulsions, consists of very fine drops of oil which dry as drops and not into a con- tinuous sheet as does ordinary oil. This may not be the explanation, however. If the reader will turn to Table No. 5 (page 68) he will discover that carbonic acid has some effect on oil. White lead paint in an atmosphere of carbonic acid was ''set" in 24 hours, but was not adherent to the wood in 72 hours. It may be that carbonic acid of carbonic acid lead unites in some way with oil. In any case, it is the car- bonic acid lead which makes the paint white and which causes it to powder ; it is the water lead which makes part of the oil into a soap and hardens the paint. Both, probably, add somewhat to the ease of working, so that we may illustrate the effect of white lead on oil by the effect of soap in washing one's hands. (See Oils.") The effect of soap-making, both in boiling oil with or without driers, and also in making a hard paint is, as we shall see, injurious to the color of the paint. Mulder recommended to artists a white lead which should contain no water lead, and therefore would iiS Painting and Painters' Materials, produce no soap ; but Messrs. AVigner & Harland, of England, have shown, as above, that this would not do as a paint. Mulder's conclusions harmonized with those of these English chemists, but were not so clear as theirs, for he was puzzled to find the effects of soap-making (ease of working of the paint) without other results produced by a soap. The Powdering of White Lead. — As we have seen, at least one-fourth of a properly made white lead is a substance (water lead) which may be compared to red lead in its effect upon the oil. The other three- fourths of the lead injure the oil as a paint, but give all the covering power, as water lead (hydrated oxide of lead) has no coloring properties whatever. The powdering of the lead is due to this three-fourths of carbonic acid lead ; the change in color to the soap- making between the water lead and the oil, producing more or less o^y-oil acid which in the shade becomes red. See Part II. We venture to make these bold statements on these disputed points because the experiments of Messrs. Wigner & Harland, of England, combined with Mulder's work and observation, throw great light on the subject. There may be another cause for the powdering of lead, as facts in our own experience indicate, and as Messrs. Harrison & Bros., of Philadelphia, state in their circular, namely, the use of too little oil with the lead. Mulder speaks of 250 to 280 per cent, of white lead as recommended by practical men ; but advises 52 parts as sufficient for all soap-making purposes for 100 parts by weight of oil. He admits, however. White Lead. up that very much more than this does not produce a hard, brittle paint, as it would do if there were much soap-making with the oil. Within the house 700 or even 1,000 parts by weight of lead to 100 parts of oil are used without producing a paint which is either brittle or powdery, and which has the great advantage of holding its color. We may, therefore, reasonably lay the blame of much, if not all, the powdering of white lead upon the carbonic acid, since Messrs. Wig- ner & Harland*s experiments prove that a pure car- bonic lead paint will entirely powder like whitewash. The hardening and quick drying of the lead is to be entirely credited to the water lead ; but if we desire a lead which will undergo the least change of color, we must seek one with at least as little as 25 per cent, of water lead. Less than this will give us a bad paint. The adulteration of white lead may be an advan- tage. Messrs. Wigner & Harland say that 5 per cent, of adulteration does not injure the covering power of the lead ; but 10 per cent. does. Aside from loss of so great covering power, an adulteration of 15 per cent, would not be a great injury to paint for rougher outside work, and if done honestly, no injury to the man who should boldly state this to be his formula of composition. White lead, as we have seen, is made by the use of acetic (vinegar) acid, and there remains more or less acetic acid lead in the product. If this is not thoroughly washed away, the spread paint has the appearance of small pox," from dry, hard masses, produced by the acetic acid lead. These will form before the lead is used. Lead made by the Dutch process is often not so I20 Painting and Painters' Materials. white as other lead, because the manure which is used to produce heat about the pots gives rise to sulphur gas, which may unite with lead and form more or less black sulphide of lead. Tan is mixed with the ma- nure, and in other countries tan and pyroligneous acid are used. The German process of making lead is in closets or chambers, not in pots ; the French process, by passing carbonic acid through a solution of acetic acid (vinegar) lead. This gives a very white lead, but one containing large particles which are more or less in the condition of crystals. Stein says the French process lead soaks up more oil, and gives this as the reason why the lead does not cover so well. The truer reason is that given on page 35. The red tinge in some white leads is apparently due to the presence of silver in the lead ore, which pre- vents some particles of lead from changing so rapidly into white lead as the other portions of the corroding metal. " In 1868 a large quantity of lead was decop- perized by means of zinc at the Sheffield Lead Works, leaving a quantity of silver, which varied from 15 dwts. to i oz. of silver per ton. About 80 tons of this lead were placed in the stocks for cor- rosion when the whole of it came down of a uniform pink tint." Some of our soft American leads contain a little bismuth, which, being lighter than lead, some- times collects at the top of the melting kettle, and buckles are produced containing much bismuth, which remain black in the carbonic-acid atmosphere where the white lead is corroded. Lead long mixed with oil (old tub lead) is much better than fresh lead ; it covers better, and has been said to change less in color, but of this fact I am not at all sure. The change White Lead, 121 is not so much in the lead as in the oil. The carbonic acid lead produces an emulsion with the oil, that is to say, it separates it in some way, possibly into very fine drops. The long standing may allow the oil time to recover partly from this effect, and it certainly causes it to unite more closely with the water lead as soap. Old lead, therefore, for both these reasons, is stiffer and does not always work so easily, but can cover more surface with a harder and more per- manent covering. Practical Use of the Above Facts.— Vi is very gen- erally believed, and on good grounds, that white lead bearing the name of even reliable firms is more in- clined to powder than was the case years ago. Com- plaints arise on every hand, and the quality of lead purchased should be, and we are inclined to believe is, an assurance that the cause is not adulteration. Lead largely adulterated with barytes might powder in this way if the adulteration were considerable enough, but we doubt that any firm in good standing sells such white lead branded with it^ firm or mill naine. Adul- terated lead is undoubtedly found under second-class brands, or under a trade title as this or that " Whiter All white lead will powder, because, as we have shown, it contains carbonic acid lead. Lead not prop- erly washed will dry in lumps or grains, because it contains acetic acid lead (which is a drier) and may also contain pure lead particles. Finally, white lead, as found in the stacks after corrosion, differs very much in composition. Corro- sions from the centre of the stack may differ very much from those at the side walls, where more mois- 122 Painting and Painters Materials. ture has been present. In some cases, with an excess of moisture, corrosions are quite sugary m appear- ance, rubbing between the fingers into small crystalline grains, which, upon analysis, will be found to be nearly pure plumbic carbonate, PbCOg" (carbonic acid lead). It is not improbable that the cause of poor quality is due to improper oi too rapid corrosion. Reference to the table of analyses on page ii6 of this chapter will show (the analyses were made by a German chemist, Weise) a less amount of water in second and third rate lead than in lead of first quality, and the inference is that these contained less water lead. But the question cannot be decided by any chemi- cal analyses thus far made, because these do not agree with each other sufficiently All that can be said in view of the clear light from Mulder's, but especially from Wigner & Harland's experiments, and from con- firmation by all the facts, is that the excessive powder- ing of any lead, not badly adulterated, is due to the small quantity of water lead it contains. What this is due to we do not positively know, except that it requires a certain time to make good lead by the Dutch process, usually from six to eight weeks. By other processes, good lead is made in shorter time ; but none of these, we believe, is successful in this country The whiteness of lead is due largely to the purity of the metallic lead used — it may also be due, as we have indicated, to the method of manufacture. The method of manufacture has much to do with the cov- ering (coloring) power. All lead which contains crystals has less covering power, and this lead, made White Lead. at least by some short modification of the Dutch process, is the one easily deficient m water lead. All the mysteries of this subject cannot be cleared up, but 1 have given the reader those conclusions which seem to me best to harmonize with and to be sup- ported by the best authorities who have made thorough investigation of this subject. Although the matter has been discussed on the ground of scientific facts, the result must be practical and not scientific certainty — that is to say, it is the best light we have to-day to act upon. CHAPTER XIII. Rust. The preservation of iron is a matter of such im- portance that the painter needs a good deal of knowl- edge about the causes of destruction which he, more than any one else, is called upon to prevent. Gold and platinum have no rust. Silver is not at all affected by the oxygen of the air, but is blackened by the sulphur gases (sulphuretted hydrogen). The sulphur in eggs rusts silver, forming the black sulphide of silver so difficult to remove from a silver spoon which has been much used in eating eggs. Copper is little affected by the oxygen of the air, but forms a green rust (carbonic acid copper, or, with acids, ver- digris, acetic acid copper, and the like). The rust of copper prevents further rust ; it forms, so to speak, a protective coat of color, usually green or bluish green. Copper rust is very poisonous. Lead. — Lead attracts and unites with the oxygen of the air, and forms a thin black rust. This does not, however, increase unless the lead be heated or exposed to hot sunlight, when yellow and even red lead may be formed. The true lead rust is white lead, and due to the action of carbonic acid in water or in the air. Soft rain-waters have much more effect than hard waters, and water containing sewage matter and fatty acids, cider and sour milk still more. Lead rust is poisonous. Tin is very little affected by oxygen of the air. It loses its lustre (as tin plate), but the acid gases of the Rust. air — especially the sulphur gases — have less effect upon It than upon silver. Red lead softens tin, but does not rust it, although some tin may be changed into a white oxide of tin. Tin plate is made by coating plate-iron (charcoal plate or coke plate) with molten tin. The protection to the iron is perfect, while the coating remains entire. When broken at any point, the iron rusts more rapidly for the presence of the tin. Water is composed of two gases linked : Water. Hydrogen. Oxygen. Water may be broken up into these gases by a gal- vanic battery. Tin and iron, or zinc and iron, form such a battery, if an acid such as carbonic acid be present. Tin and iron and the carbonic acid in rain-water act as a battery and unlink the gases of the water. Iron. Tin. Oxygen . H y drogen . * The oxygen of the water unites with the iron and, of course, rusts it more rapidly than the oxygen of the air would have done. Zinc and iron act in the same way to water, but the unlinked oxygen attacks the zinc and not the iron. The coating of rust which soon appears on bright tin plate exposed on roofs comes from the iron under- * The hydrogen may join with the nitrogen of the air to form ammonia, formed with iron rust. 126 Painting and Painters Materials, neath through minute holes in the tin. If the tin plate be examined by a glass, holding the tin so as to reflect the light, a number of minute dents or holes will be discovered in its surface through which the i'roduction of odorous oxygen (the oxygen of the air has no odor) from water containing a little acid, by means of a battery. This form of oxygen is very powerful, and is called ozone ("to smell It is its odor which is perceived by the nose during the working of an ordinary electrical machine. water reaches the iron and the rust runs out. These are more numerous on the lighter tin, and on the heavier tin fewer are probably open to the iron. They are the holes by which the paint must hold to the tin, but in case of new tin they are already filled with grease, which must come out before the linseed oil can go in. This is the reason why a roof is better for exposure for some days before painting ; the grease must either be evaporated, or it must be run out by the rain-water and the rust. The oil can then get a direct hold upon the metal inside the holes. Rust, i2r This grease comes either from the top of the bath of tin into which the iron plate is dipped to tin " it and which is placed there to prevent the air from act- ing on the molten metal, or it may be rubbed in dur- ing the polishing. Zinc and Galvanized Iron, — Zinc is slowly rusted by the oxygen of the air— more quickly by moist air, and becomes coated with a white oxide of zinc, which protects from further rusting. Galvanized iron has the same disadvantage as tinned iron: it acts as a battery to unlink the oxygen of rain- water. The oxygen, however, attacks the zinc, while on tinned iron it attacks and rusts the iron. It is better that the plates of corrugated iron should be galvanized after they are bent into waves, because this bending into shape is apt to crack the zinc. At least it is an advantage to get plates which have been regalvanized after bending, to fill up cracks. Zinc of galvanized iron is quickly acted upon by sea air, and by all acid air. Soot is very destructive to it, soot, zinc and the carbonic acid of rain-water forming a battery like those described above. Soot is a compound of ammonia ; the urine and dung of animals contain acid and become ammonia. Therefore all these are destructive to zinc roofs, and cats especially should avoid such hunting-grounds. Lime-water and water from oak wood are also injuri- ous to zinc. The Chemistry of Iron Rust. — The rust of iron is always a linking together of the oxygen of the air with the iron. And yet pure dry air has little effect on dry iron. White vitriol (sulphuric acid) absorbs water very rapidly from air. In a bottle containing white 128 Fainting and Painters Materials. vitriol a briglit iron wire may be hung from the stop- per (air-tight stopper) and will, one may say, never rust so long as it does not touch the acid. On the other hand, water without air does not rust iron. A piece of iron shut up in a sealed glass tube containing water whose air has been driven out by boiling, never rusts, if the conditions above are properly fulfilled. By rust red iron rust is intended — not simply a dull- ing of the lustre of bright iron. Bright tools may be kept for a long time in water in which carbonate of soda remains dissolved. We may say, therefore: (i). Air alone does not rust iron. {2). Pure water without air does not rust iron. (3). Air and water alone rust iron slowly. What is needed for rapid rusting ? The presence of some acid. Carbonic acid does exceedingly well. Professor Calvert, of England, made a very long study of the rusting of iron. The following were the results of experiments lasting a considerable time : Dry pxygen and iron No rust. Damp In three experiments only slightly rusted. Dry carbonic acid and iron No rust. Damp " A little white carbonate of iron. Dry carbonic acid and oxygen and iron No rust. Damp carbonic acid and oxygen andiron Very rapid rusting. Dry and damp oxygen and am- monia and iron No rust. For rapid rusting, therefore, we need (i) air, (2) water, (3), carbonic or some other acid. The last is always present in the air, and rain brings down a good deal of it dissolved in the drops of rain-water. Coal contains sulphur, and burned coal gives sul- phur gas ; coal cinders with water give some sulphuric acid. But the burning of coal especially gives large Rust. quantities of carbonic acid gas. We have, then, on the line of railroads and in cities a special cause of rusting in iron, due to the presence of carbonic acid from coal fires. An examination, made at the Stevens' Institute some years ago, of rusted iron from two bridges on the Pennsylvania Railroad, showed in water from the iron, sulphuric, a trace of sulphurous, and a quantity of carbonic acid, besides chlorine; and the conclusion reached, after some experiments, was that the pres- ence of carbonic, sulphuric and sulphurous acid was sufficient to cause corrosion, as, indeed, the over-head work of abridge subject to smoke will show. One of the over-head bridges, or, rather, over-head coverings (galvanized iron), of one of the New York elevated roads was found to need renewal twice in a few months. Still, there is a more efficient agent for producing rust than even carbonic acid, to wit, rust itself. All rust (not having been roasted) contains water, es- pecially all yellow rust. The rusting of iron seems to take place after this fashion : Air. From which iron takes oxygen and becomes : Rust No. I. — Iron and oxygen. This process goes on until we have a combination like the following : 1. Iron and oxygen. ) 2. Iron and oxygen. > Oxygen. ) This is our common red rust, or, if there is much water in it, yellowish rust. Paintmg and Paiiitei's' Materials, Now the iron underneath this rusty coat steals the last oxygen and becomes by the same process : Again we have red rust. To make up for this loss of oxygen, the upper crust has been taking more oxygen from the air. And so the process continues, because the first oxide of iron (iron and oxygen) has a great liking for oxygen, and because the red rust easily gives up some of its rust to its neighboring particles just below, while helping itself to more oxygen from the air. This process of rusting (rust causing rust) is quite peculiar to iron. Other rusts, on zinc and lead for example, do not give up their oxygen to the metal underneath, but remain To this peculiarity iron rust owes its power of spreading under tin, zinc or paint used as a protec- tion. There is one kind of iron rust (iron and oxygen) which does not spread in this manner, namely, the magnetic, black oxide, which is found on Russia iron and is produced by the Barff process. Apparently the black scale (iron and oxygen) found on wrought and cast iron, falls off because of the form- ation of other rust under it. A distinction must, however, be made between this scale and the hard skin of sand, etc., found upon cast iron. If this be painted over immediately, or dipped into hot oil, it Zinc and oxygen. or Lead and oxygen. Protection against Rust. retains its hold, while the scale on wrought iron con- tinues to fall off, after being covered with paint. The whole difficulty in protecting iron by paint lies in the fact that no covering of paint of any kind will protect the iron from rusting by the action of its rusty either red rust or common black rust. There is no paint which will kill" the rust. Paint can only water-proof iron. THE RELATIONS OF PARTS OF AN IRON STRUCTURE. The questions involved in the influence of one part of an iron structure upon another as regards rusting are too important to be passed over. Engineer Clark, of the Britannia Bridge, observed that a pile of plates, in all respects the same as those in the bridge, had rusted so badly that they were swept away with a broom, the bridge plates bolted together in the bridge suffering meanwhile no corrosion. This led him to the conclusion that iron might be protect- ive of iron through connection, and to test the suppo- sition he placed two plates on the bridge, one bolted to it, the other insulated by glass. The insulated plate evidently rusted the most rapidly, but the exper- iment was brought to a close by an accident before time enough had elapsed for a thorough investigation. In direct contradiction to this, Mulder gives it, as the result of observation, that small pieces of iron con- necting larger ones suffer so severely by rusting that it is better to paint the small ones thoroughly and leave the larger ones bare rather than vice versa,^ The greater the difference between the size of the pieces the greater galvanic effect. He recommends, there- * He says if a platina plate be thus connected with two pieces of iron, the galvanometer shows a current is flowing. TJ2 Painting and Painters' Materials. fore, the dipping of nails and screws into coal tar in order to keep out all water and damp air which the galvanic action would quickly tend to decompose. Here we see the electrical theory working in exactly opposite directions in the hands of two of the most trustworthy scientific men. A recent experiment of Farquason's in Portsmouth Harbor shows that both may be right, and is, more- over, an evidence of the care needed in studying these questions. Plates of iron and steel (each kind as nearly alike as possible) of the same size were placed under sea- water, couples of iron and steel bemg connected and other pieces left isolated. All remained in the water for six months, and when weighed each had suffered loss by corrosion as follows: i Loss. ^ Oz. Gr. ^,^^^H connected] ^ iron ) \ 7 417 Steel 3 340 Iron 3 324 f^f [connected ];:;;.::::;;;:::.•:::;:;;;;;;:::;::;::::• ° Steel 4 Iron 3 190 f;„-'[connected^ % 337 Steel 4 157 Iron 4 57 The three iron plates connected with steel lost 21 57 Isolated plates... 11 137 The three steel plates connected with iron 4 187 Isolated plates 12 60 Connected plates iron and steel 25 244 Isolated 23 197 It will be observed the total rusting is increased by contact about Q per cent. The iron lost by contact with steel about (excess) 100 The isolated plates of iron and steel lose each about tl>e same. This would have been still more instructive if iron could also have been connected with iron and steel with steel in large masses, but as it is, no experiment made upon the rusting of iron is more suggestive. Protection against Rust, 133 for, as Mr. Farquason suggests, and as has been shown by previous experiments, using polished sur- faces, iron rusts unequally, and it is probable that one part of the mass, harder than the other, will be de- structive of the whole. Mallet, who many years ago investigated the rusting of different sorts of cast iron, reached the conclusion that chilled iron rusts most rapidly, and that the more homogeneous and closer grained and less graphitic the iron the less the corrosion. We see, therefore, that it is quite certain that differ- ences in hardness and in structure of different pieces of iron connected may result in the greater corrosion of one than the other. Also that Mulder may be quite right when he says that it is better to paint the small connecting pieces and let the larger pieces alone than vice versa. As to the other conclusion, that iron connected with iron is protective, there appears to be no certain evi- dence aside from the effects of the vibration to which such a piece may thus be subjected. It has been pointed out that a pile of rails or the rails in a side track rust much more quickly than rails in the main line. But all the causes of this difference have not been demonstrated. The first question to be settled is the amount of water remaining on each kind, and the amount of rust, also the mere physical effect of vibra- tion as related to these. Certain only is the fact that vibration is protective against rust. Connection with other Metals. — The question of con- nection with other metals is an important one in rela- tion to the preservation of iron, and the question de- 134 Painting and Painters* Materials, mands a closer study than we are here able to give it, for want of a complete set of facts. Copper \w connection with iron is very destructive; on this all who have investigated the matter are agreed. Professor Colton says that care should be taken that no copper comes into connection with ship iron. Zinc. — Zinc by itself is one of the most durable of roofing materials. Prof. Max Pettenkofer was at the head of a commission which investigated the decay of zinc, and experiments on a zinc roof 27 years old showed that the rate of loss would destroy a roof one- fourth of a line in thickness in 243 years. At a meeting of the Society of Engineers (English) many years ago, the following was given as the ages of zinc roofs (Belgian zinc) still in good preservation. With galvanized iron the case is different ; it is val- uable only when the conditions and its manufac- ture are such as to keep a perfect surface of zinc. Clark condemns its use in all acid atmospheres, but advises its use elsewhere, and there is a mass of testimony to confirm this advice, but it must not be forgotten that the question of the thickness and per- fection of the zinc covering is involved in any such conclusion. Under water, experience with galvanized iron has been unfavorable. At first, zinc was supposed to be valuable as a protector to iron ; but it was found Years. The Cloisters at Canterbury Portsmouth dock-yards Great Western Railway Station at Rugby Another railroad station 33 24 ^5 Zinc and Paint, that the protection was only partial, and not lasting. In some water gates in which zinc nuts were screwed over the iron bolts to prevent corrosion, the iron was attacked after three years. Nevertheless, above ground inquiry into the comparative life of galvan- ized and common telegraph wires showed the life of the bare wire to be 15 or 20 years, while gal- vanized wire of 20 years' age was found to be but still little worn. It may be said, therefore, that zinc-covered iron is valuable in proportion to its perfect condition, and for those conditions in which this perfection of cov- ering can be kept ; an acid atmosphere is destructive as is also sea water. Zinc and Paint, — The difficulty of making paint stick to zinc is, I think, of a different kind from that ordinarily supposed. Boettiger, who has studied the question somewhat, recommends the following as a valuable wash. Its purpose is to change the metallic zinc surface into zinc chloride and amorphous brass : I part chloride of copper. I " nitrate " I " sal ammoniac. 64 water. The zinc surface is to be washed with this* and left for 24 hours, which will give a black surface on which one can paint, but which will probably shell off the iron in the liveliest manner when any spot breaks up. Mallet found that zinc added to copper in connection with iron increased its rusting 60 per cent., cop- per without zinc only 40 per cent. A better method ♦Wash thoroughly with pure water afterward. Ij6 Fainting and Painters' Materials. is probably simply washing the surface with dilute muriatic acid. This, however, will result in white lead turning as yellow as you please, as a Boston man discovered who used this preparation on a zinc ceiling ornament. What we need to know is, why the oil does not stick. This, however, we do not certainly know, and can only guess at. One reason is because the surface is very smooth, and when the oil dries up, as it soon does, the paint having nothing to hold by, peels off. Another reason appears to be that the zinc does not unite with the oil acids, and, in fact, exercises some injurious action upon them. We should recommend, however, a firm paint, such as given by mixing some lead with iron oxide.* In conclusion, it may be said that the great value of both tin and zinc-covered iron plates lies in the fact that only in this manner can we place the plates in position free from rust. If we were able to place naked iron in the same positions as thoroughly free from rust, it is not impossible it would be better so to do, because paint will hold to such iron better than to zinc or tin, and we should avoid all galvanic action. The experiment has been tried of finishing plates simply painted, but the paint rubs off in transit, and any bare spot may endanger the whole plate. Finally, it may be said that except in sea water, there is no evidence of injurious galvanic action be- tween iron surfaces and metals used with and covered by oil as paint. ♦The experiment may also be made of adding a little not-drying oil to off- set the excessive loss of these oil acids, which keep the paint layer pliable and prevent over-drying. CHAPTER XIV. Painting Iron. It IS plain, therefore, by the facts stated in the last chapter, that we must discriminate in painting metal as to what we shall do in the way of preparation. It will also become evident that it is well to discriminate in the choice of our paint. Cast Iron. — Cast iron does not rust rapidly, but in salt water softens and become a sort of plumbago. It should be painted or dipped in hot oil as soon as it leaves the mold, in order to preserve the hard skin before it is acted upon by the air. If rusted in the least, the rust must be removed by scraping or brushing with wire brushes. Wrought Iron. — The black scale or rust on wrought iron must be removed. It is not necessary to remove all the black scale, and indeed it is a very difficult thing to do. Mulder (who investigated the painting of iron for the Dutch government) found that a piece of iron from which he had with great difficulty re- moved this scale rusted quickly, while the piece on which the black scale had been left was quite free from rust. One will often observe pieces of old wrought iron quite clear of rust from no other protection than this scale. We know that the scale itself is of different kinds ; that produced by superheated steam, for example, being, as we have said, the best and only sure protec- tion, while the ordinary scale to a large percentage ij8 Painting and Painters* Materials. gradually falls off, from (apparently) the formation of a minute quantity of rust powder under it. It is worth while, therefore, to study the varieties of this (black) scale, for there is without doubt a differ- ence between that formed in the ordinary conditions of the air, in water, under the beating hammers, etc. This study we must, however, leave the reader to make for himself, warning him only that all ordinary scale (black) is of an unreliable character, in the fact, apparently, that it can give up oxygen to the iron beneath it, and so form crumbling rust of the stuff by which it is held to its bed. It may be painted, and the paint will hold fairly well for a time, if undis- turbed, yet, when the paint is scraped the scale will come with it ; and this is a method sometimes used to get rid of it on ship iron. A light coat of paint is put on, which, after a year, is scraped off, carrying the scale with it, and leaving a solid surface for perma- nent painting. The scale may be removed in several other ways. (i.) It may be allowed to rust off, and the iron may then be scraped with wire brushes. The objection to this method is the same as that against allowing your garden to get full of grown weeds in order to pull them with ease. In either case your antagonist may prove too strongly planted. (2). The iron may be pickled in a bath of i part of sulphuric acid to 100 of water ; or washed with a solution of 8 parts of acid in 100 of water; after which well scraped with wire brushes, and rewashed with lime water. Acid is only allowable where it can be thoroughly neutralized with lime water. It is the practice of the Cincinnati Southern Rail- Painting Iron. 139 road, to whose experience we shall refer, to have its iron treated with hot oil as manufactured without mix- ture with pigment of any kind, no paint being used for some months after the erection of the bridge ; and then a coating of red lead. When no attempt is made to remove the rust or scale, these must be mixed with the oil in putting on the paint, as well as that can be done with a brush. The object is simply to coat the rust with oil so that it will not be in contact with the iron, and is effective of course only with the loose particles. Steel. — Recent English experiments^n steel (using a galvanometer) give evidence that the pitting " of steel is partly due to the galvanic action between the black scale and the body of the steel. I have not seen the records of the original investi- gation, but Mr. White, of the Admiralty office {^Jour- nal of the Iron and Steel Institute^ 1881) says : That was not a speculative belief, but a belief based upon experience and many careful experiments made under water in Portsmouth harbor. The trials were made with the greatest care under the most varied conditions, and the results made it as certain as one would be certain about anything that the black oxide, if left on portions of steel plates, would cause pitting on the bared surface of the plates. Active gal- vanic action could be traced with the galvanometer on the parts of the plates from which the scale had been removed. There was all the difference in the world between corrosion and pitting * * * miralty experiments proved. Mr. Parker had shown that there might be practically no corrosion on clean surfaces (steel) during very long periods, but if a hole was formed in the plate by pitting, it became a serious I^O Painting and Painters Materials. matter. In the Royal Navy they were trying to get rid of the black oxide by means of pickling before being worked; * * * plates were also being dealt with by a * * * process which had not been per- fected. * * * In the private trade, where ships were built in the open air and exposed to the weather, he believed there was less difficulty in getting rid of this scale than in the case of the Iris and the Mer- cury. Great care had been taken * *, but when the Iris had been on service in the Mediterranean a few months, it was fc)und that the effects of the scale were visible." In the paper referred to by Mr. White, Mr. Parker says : We have, at the present time, i,ioo marine steel boilers running, * * * ^svd the accounts I have received down to the latest go to show that steel boil- ers behave in respect to corrosion about as well as iron boilers. Greater irregularity in the corrosion of the steel is reported, and I am inclined to the belief that this is due even to a greater extent than in my experiments to the unequal action of the scale, and if it should be found necessary to remove the scale, the difficulties in the way would not be great, and much irregularity and pitting would doubtless be removed." Priming Iron. — Having rid the iron of all but fast- hard rust, the question of a priming arises. Nothing can be done by any paint except water-proofing. Paint can have no influence on iron except to rust it by acids, or else to keep its surface free of water. The chief thing is to get the water-proof er to stick, (i) On smooth iron surfaces, oil loses its soft, not-drying acids and dries up ; on rougher surfaces, it holds to Painting Iron, iron quite as well as to wood, but (2) on all surfaces with black scale, the scale may leave the iron and carry the paint with it. Besides, there are only two other difficulties in painting iron. (3) The rust may spread under the paint from any point. (4) The paint which will stick to the iron best will most readily crack open the coats above it. These are the chief difficulties, but Dr. Wiederhold maintains after a considerable study of iron surfaces that there is another one, which he apparently regards as the principal one. He claims that unless the priming color dries quickly, or if turpentine is used, the cooling of the iron (as at night or by the rapid evap- oration of the turpentine) causes a deposit of dew-like moisture upon the paint. This moisture forms an * 'emulsion" with the paint, and it does not dry into a homogeneous mass. His idea seems to be that paint drying under the influence of water on its surface leaves little holes to the iron, or else that it more easily peels. I am unable to give the value of this result, but do not, my- self, have any great confidence in it, but give it as the conclusion of one who claims to have made a study of the subject. I have not observed the rings (caused by drops of rain on fresh paint) to be centres of rusting. Taking up the last difficulty (4) first, we find that cracking of paint on iron and the wrinkling and roughing up of paint come mainly from too thick a coat. Nothing will stick to iron better than boiled oil, because it is when dry, {a) firm, but (c) soft under- neath, (^) elastic, so that it holds together well. The objection to it, however, is that it is too thick— it leaves so thick a covering that it wrinkles it, and may crack any over coat. Painting Cast Iron. — Cast iron usually has such a rough surface and so many pores that there is no great difficulty in making paint stick to it, if the rust does not get planted underneath. When such an accident 142 Painting and Painters^ Materials. occurs the only remedy is to scrape off the paint and rust at that point, and repaint the spot. It seems to the writer that it is for want of this dis- tinction between the condition of the surfaces of irons (cast and wrought iron) that there is some confusion of views about methods of protecting them. Both have this difficulty, which no paint of any kind can overcome, that the rust spreads from any point under the paint. On both (exposed to the sun) the hot iron and the sun combined drive off the not- drying oil acids which keep the dry paint soft ; the paint layer becomes hard and brittle. Here, however, the two surfaces differ. The cast iron has a rough surface, and if promptly and properly painted is not inclined to scale. The scaling of the black oxide from the wrought iron carries the paint with it. Wrought iron needs a paint hard and elastic, which will hold itself together even if points of scale give away underneath it : hence the value of red lead on wrought iron j while on cast iron, other paints, iron oxides for example, will serve quite as well. The fol- lowing experiment was made under the auspices of the Dutch State Railroads. Iron plates were prepared for painting, as fol- lows : Sixteen plates (Nos. i to 16), pickled in acid (hydro- chloric), then neutralized with lime (slacked), rinsed in hot water, and while warm rubbed with oil. Sixteen plates (Nos. 17 to 32), were cleared of scale (so far as it could be removed) by brushing and scrap- ing. Four plates from each set were then painted alike, for example, Nos. i to 4 and 17 to 20, with coal tar ; Nos. 5 to 9 and 20 to 24 with iron oxide A ; Nos. Painting Iron, T43 lo to 13 and 24 to 28 with iron oxide B ; and Nos. 14 to 16 and 28 to 32 with red lead. They were then exposed for three years with the following results : Paint, On scrubbed plates. On pickled plates. Coal tar. Quite gone. Inferior to the others. Iron oxide A* Inferior to other two. Holds well. * ' B Superior to A ; inferior to red lead. ** Red lead. Equally well on both, and superior to all. It is seldom that v/e can obtain so clear a record covering so long a period as this ; but how shall we explain it ; what, in other words, are the differences be- tween the surfaces and between the paints ? Pickling takes off quite all of the black oxide ; scrubbing does not. Red lead unites with oil to form a hard oxy- linseed-oil-acid soap — a harder soap than given by any other combination. Says Mulder, of boiled oil and iron oxide, boiled oil and litharge, boiled oil and zinc white, and boiled oil and red lead : " The lead, when dry, contained the largest amount of oxy-linseed-oil- acid lead, and was harder than the others." It seems to me, therefore, that the best interpreta- tion of the facts is the simplest, namely, the difference in the plates is the scaling off of the black oxide on the scrubbed plates ; the difference in the paints, that the red lead did not give way when this scaling oc- curred, because it is a hard, elastic layer of paint, which holds itself together by its own cohesion. Further Testimony, -This, result in the above experi- ment was not, however, exceptional. The testimony to red lead is almost universal. ♦From an analysis of this paint, it is probable that it contained more clay than the other. 144 Painting and Painters' Materials. The Cincinnati Southern Railroad has a number of miles of iron trestle work and bridges and has there- fore furnished a large experience in the use of paint as protection for iron, the results of which may be summed up as follows : Red lead has proved by far the best and most last- ing paint. It is difficult to obtain in purity and is more expensive than iron oxide, but so much more lasting in the climate of the region through which the road passes that it is used in preference. The iron oxide is washed away by the rain and perishes in spots, endangering the iron ; and yet nevertheless is a valu- able paint if frequently renewed. The standard paint for bridges on the Pennsylvania Railroad is white lead and a little yellow ochre, over- heads being coated with asphaltum on account of the color and greater durability under the effects of the smoke. Prof. Henry L. Colton writes to the Scieiitijic American that after three years' experience, with un- limited resources for experiment at his disposal, he finds nothing equal to red lead for ship iron. An instance comes from England of pump-rods, in a well 200 feet deep, which had stood forty-five years painted with red lead. At the expiration of that time their weight was found to be precisely the same as when new — no loss by rust. Nearly every country of the world can furnish tes- timony of the same kind ; but some people must have a thing happen in their own backyard with nobody by to make an exact record, or else, it's all theory." Disadvantages of Red Lead, — Red lead is adulter- ated with brick dust and other substances, and in this Red Lead Paint on Iron. 145 way has lost, perhaps, some of its good reputation."^ Its value is that it unites with the oil, giving up at the same time a part of its oxygen. No other substance which does not unite with oil can replace it. Its value is its effect on the oil acids. Partly on account of this effect, it is said at times to blister badly. Spon says that should chemical action commence, red lead blisters, and is reduced to metallic powder. This is possible,f and it is also probable that red lead under great heat will blister sooner than iron oxide paint ; and part of the reason is plain : it is more elastic. Elasticity in paint, as we shall show in another arti- cle, comes from the not-dried-up oil acids. Red lead has also been accused of forming a bat- tery with iron, and rusting the iron faster by unlinking the oxygen in water (see last article). No evidence within my knowledge as to the effects of dried red lead paint upon iron has appeared except as to the effects upon two vessels, and on these below their water line. The paint on the hulls of the ships William Fairbairn " and ''Guienne," painted with red lead, was found (under the water line) blistered, and in these blisters were metallic lead and chloride of iron from the effects of the battery made by the lead and iron on the sea water. But Jouvin, who reports the case of the " Guienne," says that above the reach of the sea water wherever the red lead was in good condition it had done no injury. In fact, we cannot be sure that even in this * In Europe it is also adulterated with colcotha (English red), made from refuse iron in the manufacture of sulphuric acid, and therefore quite cer- tainly contains sulphuric acid, which would of course be very injurious to the painted iron. •j* A master ship painter of very large experience tells me that blistering' does not take place unless the vessel is painted while wet, or its iron more or less covered by scale. 1^6 Painting and Painters Materials. case all the trouble may not have come from painting the vessel while wet. Our government vessels are painted with red lead and zinc ; but the favorite paint for ships' bottoms in the merchant marine is, I be- lieve, red lead alone. Red lead gives with zinc a very hard paint. Red lead softens tin, and has been ac- cused of eating holes in it ; this is not probable. It should not, however, perhaps, be used on tin. The tin of a cup containing red lead paint can be easily scraped with a knife, its surface coming off in thin shavings." Finally, the color of red lead is not durable, espe- cially not with white lead. Under the action of the sun it becomes less orange ; and mixed with other tints, or under the influence of sulphur gas, its color is fugitive. Again, red lead is perhaps more injurious to the workman than other paints, because he must mix it with oil as it will not keep for use ground with oil. The safety of the workman against the poisons which he uses lies in cleanliness and ready-mixed paints. The iron oxides should be preferred where they will serve equally well. COAL-TAR AND ASPHALTUM. Coal' Tar is largely produced in the manufacture of illuminating gas, the crude flying constituents of coal being driven off by heat and passed through water where the oily impurities of the gas are con- densed as coal-tar, the partly purified gas flying away. Coal-tar distilled gives light oil,* dead oil and pitch. ^ Pitch gives artificial asphalt. * Light oil is divided into first light oil" and " second light oil," which together form crude naphtha, which redistilled gives commercial benzol. Mulder's benzol is probably crude naphtha. Asphalt and Coal- Tar. 14"/ Asphalt^ as a natural product, is a limestone im- pregnated with bitumen or asphaltum, which resem- bles coal-tar. When speaking of asphaltum, one may signify either (i) the rock in its original state, or as mixed and prepared with other substances as a paving material or cement ; (2) or the 6 or 7 parts of asphal- tum obtained by boiling the limestone ; or (3) coal- tar pitch in either its crude or its purified state or mixed with other substances, or the coal-tar itself mixed with lime and sulphur, etc. Altogether there is a delightful state of uncertainty about the word. Practically the difference between the crude asphalt (rock) and any manufactured article appears to be the better incorporation of the resinous and oily bitu- men with the limestone, so that it is less brittle and more elastic. In addition to this the true asphalt (distilled) appears to hold fast an oily substance which escapes from the manufactured coal-tar. The natural asphalt has a more glossy fractured surface, and as cement or paint does not so quickly soften under heat. Coal-tar is very brittle at the freezing point of water, true asphalt is tough at 10 degrees lower ; coal-tar softens at 115'' F.; true asphalt does not soften at 170°. Coal-Tar Paints, — A chemist who has studied asphalt has likened it to amber, and for practical pur- poses we may regard both coal-tar and asphalt (2) as consisting of a resin and an oil. Mulder experimented with various preparations of coal-tar, and the following are the results of its dry- ing as shown by weight. No. i is a coal-tar com- pound of the more resinous part (remaining after 148 Painting and Painters Mate/Hals. distilling off one-half the substance) dissolved in the lighter oil (number one). No. I. No. 2. No. 3. No. 4 Date. ■ 9.P r- p . < 9 2. 00* ; p 3 p ^9 N r+ 2- ^ • P • 3 oc\ A wv\ 1 1 . 248 J • J, . 707 2 • 323 22 " 23 " 26 " .... 30 " .... In 10 days 26 July. In 87 days Lost. 0.293 0.003 0.015 0.002 25 p.C. .028 2 p.C. Lost. 0.325 0.045 0. 116 0.066 14 p.C. .209 5 p.C. Lost. 0.462 0.025 0.056 0.042 21 p.C. .157 5 p.C. Lost. 0.363 . 040 0.061 0.042 22 p.C. .133 5 p.C. It will be seen from the above that these varnishes are constantly losing in weight — the above experi- ence was not in the sun, where the loss would proba- bly have been greater. No. I and best produces a harder and more brittle varnish and continues to lose substance — two per cent, in weight in 87 days. No. 2 is much longer in drying and is not hard after two weeks. No. 3 much the same as No. 2, and No. 4 is even less dry after two weeks. Practical Experience. — The difficulty with coal-tar asphaltum as paint is that when hard it is brittle and when not too hard }t has a tendency to flow, and will not therefore stand either the heat of the sun nor the diffused heat of summer. So long as it will stay, how- ever, as a perfect covering it is perhaps the best, or Coal- Tar on Iron. 149 one of the very best, of paints for iron. If put on hot, it enters all the pores" of the iron and from its nature as a loose fluid-like compound keeps an imme- diate contact with every point of the surface. On locomotive smoke-stacks it dries hard, and, occasion- ally, bits of it blow into the eyes of the engineer. Even on underground pipes it has a tendency to flow, and for this reason, as shown by the experiment in above, it is probably less reliable for iron than iron oxide paint. It has its value, however, and may be said to be still on trial. Mulder recommends it as perhaps the best protec- tion for iron if not put on too thick, but his climate is not ours; and even in his climate, as we have shown, it proved on a fair trial less lasting than red lead or iron oxide. An engineer of large experience with bridges tells us he regards asphaltum as a lasting paint, but as manufactured has proved to be brittle, tending to scale. It is much better adapted for damp places than oil paint, because the air of such places often contains ammonia, which makes hard oil-leather fluid. For this reason alone, and for cheapness, it is well fitted for a covering for pipes, and it is the best paint for iron under water. It should for this and all other purposes be applied, if possible, to the iron while it (the iron) is at a red heat. If this cannot be done the asphaltum may be heated, but this method is less reliable unless the iron can be dipped in the hot fluid. Dr. Angus Smith's method for iron pipes is well known ; the pipes are cleared and heated to 700** and i^O Painting and Painters' Materials, then dipped into a mixture of coal-tar pitch contain- ing 5 or 6 per cent of linseed oil. Matthewson (Works in Iron) recommends it, under ordinary oil paint, if so desired, in the following for- mula : 30 gallons of coal-tar, fresh, with all its naphtha retained ; 6 lbs. tallow ; i Yz lbs. of some resin ; 3 lbs. lamp-black ; 30 lbs. fresh slacked lime finely sifted, mixed immediately and applied hot. He claims a special advantage for the paint in the fact that, being black, any worn spots on a white or light coat of oil paint over it would soon be perceived. Its tendency as an undercoat would evidently be, however, to crack its oil leather over-coat. CHAPTER XV. Paints for Iron and for General Use Out of Doors. Although we have already entered into details as to the qualities of paints, considerable remains to be said about their use out of doors, especially as re- gards protection of iron, as well as regards their use in finer and decorative painting, which will be fully treated of in part second. White Lead. — White lead is so much used for paint- ing iron that we should be able to give some definite experience in regard to it ; yet we cannot. It is sup- posed by many painters to rust iron, and yet one of the most observing painters that we know is of the opinion that (judging from experience) the chances are even that the greater appearance of rust under white lead on iron is due to the greater contrast between the rust color and the white lead. Theoretically, of course, as white lead contains car- bonic acid, it is able to rust iron rapidly, on the sup- position that the carbonic acid separates itself from the lead. This it is supposed to do when, on painting a black surface with white lead, the black surface after a time shows through the lead color much more plainly than at first. All that can be said, so far as any facts are con- cerned, is that white lead should not, if possible, be used in priming iron, nor in any priming coat. More- over, it is a less desirable over-coat than iron oxide. It probably, however, equals zinc in durability, and ex- 1^2 Paijiting and Painters' Materials. ceeds it wherever, as on railroad bridges for example, the paint is exposed to acid gases. Zinc is very quickly affected by acids, as is also the white lead as a com- pound of carbonic acid, the carbonic being driven off and the whiteness of the paint as well. Stevens, the very best authority on house painting, writing after years of experience, recommended its use for town and city houses because of its stability of color. Pure white lead and pure zinc are here referred to, while as a protective coat for iron neither should be used pure. The powdering of white lead has already been discussed, and the experimental evidence as to its cause gets the strongest practical proof from out- door use of white paint. Nearly anything mixed with white lead will greatly improve its durability. Even a very minute percentage of Indian red and lampblack shows its effect plainly when we compare the dura- bility of the pure white paint on the clapboard and the mixed paint on the moldings of a house. The one will generally be found after a couple of years (ac- cording to the exposure to the sun) to be in the condi- tion of a whitewash, while the other and hotter color will powder off but little. Unmixed white lead should never be used as outside protection, unless its color is very much desired. Something should be mixed with the lead to pre- vent the effect produced by the carbonic acid lead on the oil. The oil is emulsified (separated) in some way, as into finer drops, and the mixture of al- most everything with the white lead tends, apparently, to overcome this effect, whatever its cause. The writer does not pretend to be able to say whether this White Lead. 153 change is chemical (change in the oil acids), or physical (change in the condition of the oil). Either change might account for the facts. If a mixture of white lead and oil dried on glass, or a mixture of white lead and oil be examined under the microscope, using a high power and passing the reflected direct sunlight through the layer, an optical effect is produced as though the portions not cov- ered by the lead, but centering around each separate particle, con- sisted of minute drops. Working with direct sun-rays and through glass is productive of optical effects which are uncertain as to their cause, but the appearances in this case are very distinct and may occur under the same condition in other pigments, in which case the size of drops should be measured. The precise effect of the lead on the oil, however, is of less consequence. The facts in the case are: (1) that pure carbonic acid lead unmixed with the oxy- hydrate produces with oil an emulsion and not a paint ; (2) that only a mixture of 25 per cent, of oxy- hydrate lead brings it to a drying-hard condition ; (3) and, although our white lead usually contains more oxy-hydrate than this, it is still improved by sub- stancds which do not affect the oil as does the carbonic acid lead. Finally, the oxy-hydrate tested alone unites with the oil with a rise in temperature, show- ing true chemical action. White lead is not a whitewash, only because it con- tains a little water-lead (oxy-hydrate) which chemi- cally unites with the oil. It is probable, therefore, that - by increasing the amount of soap (water-lead acts by forming a soap) we still further improve the lead, although necessarily by loss of its purity of color. If a white color is a necessity, a mixture of zinc white and white lead is to be recommended, but it will be well to use a little drier with the paint. If purity of color is essential, care must be taken that the zinc 1^4 Painting and Painters' Materials. contain no sulphuric acid, as this will attack the lead ; zinc from ores containing sulphur combinations con- tains small quantities of sulphuric acid. Almost anything will improve the carbonic acid lead. Even the lower grades of white lead, which, as we have shown, contain a good deal of barytes, zinc and lime, are thought by some to stand quite as well as pure lead. It is hardly necessary to say, how- ever, that no such combination is to be recom- mended * except to illustrate the fact that almost any- thing with pure white lead is an advantage, wherever the paint is to be exposed to the sun, The weather" alone has no very injurious effect on the lead unless the sun gets at the paint, but after the sun has shone for some months the rain and storm can beat off the lead as though it were only lime. It shows the need of thorough investigation of these questions that this effect has been accepted by painters of considerable chemical knowledge as the result of the chemical combination of the lead with the oil, while it is only because of this combination by less than one-third of the lead that the paint holds at all.f White lead is graded according to covering power, whiteness and fine quality ; we cannot say positively, that a first grade contains more water lead than the others, but this is probably the case, and second and ♦Rose has shown that sulphate of barium (barytes) is acted upon little by- little by water, sulphuric hydrogen gas, and possibly sulpuric acid, separated. Hunefeld has shown that sulphate of lime is decomposable under ordi- nary circumstances by the galvanic action of two metals in contact. Both barytes and lime, therefore, can give off acid under the conditions to which paint is exposed. + For these facts we are indebted to Messrs. Wigner & Harland, of Eng- land, who have done more recent work on white lead tha.i any other chemists. Iron Paints, 155 third grade leads are not to be relied upon, even if the lead has been properly corroded, they may be adul- terated. By this, we mean that they may not give so long a service. We have no experience with different grades to record ; indeed, an experience of length of service, which is definite and of value with varieties of paint, is much more rare than should be the case. Iron Paints. — These are of two classes and many kinds. In the first class are the artificial iron paints, which are commonly made from copperas and from refuse iron from the manufactories of sulphuric acid, dyes, etc. These contain more or less sulphuric acid and are never to be used on iron as a first coat, nor as a second coat if it can be avoided. They are of higher price and in a variety of colors, and only as colors would there be any temptation to use them. The second class of iron colors consists of ores of natural iron rust combined with clay or with some forms of silica. Their value for iron depends upon the fact that they contain no water and no sulphuric acid. If there is water in these rusts (paints) they will act upon the iron as common rust. From 70 to 50, and even less, per cent, probably consists of oxide of iron, the remaining 30 to 50 per cent, being of clay and other substances. The question of this substance, therefore, be- comes important. Clay has a strong liking for water, and if not soaked with oil will possibly take up and hold a load of moisture. On roofs this is of less consequence, since tin is not itself rusted by such moisture. Here plenty of oil will be used, which the clay will soak up (ochres soak up 75 or 80 per cent.), and this oil will be protected from the sun and the air. Probably for this reason such pigments as Spanish brown are so admirable for roofs. On iron, however (perpendicular surface), less oil will be used, 1^6 Painting and Painters Materials. and the clay, less well protected by oil from water, will be in dan- ger of giving- our water-proof layer a moist character. Unfor- tunately, there seems to be no exact observation, as there is no ex- act knowledge of the composition of such paints. *' What is generally called iron oxide is generally a clay (or chalk) colored with — to — per cent, of iron oxide, and ranging from yellow to brown, in color, and in weight accordingly, the lighter shades being unaffected by the magnet. With this inferior material the oil is liable to be absorbed from the iron oxide and the whole to change color and peel off. But * magnetic paint,' or pure iron oxide, is excellent as a protection for iron, it being sometimes impossible to scrape it off ; excellent on woodwork. It is resistant of salt water. It stands sulphurous gases well." Moreover, this dry and roasted rust (paint) does not act as a protection to the iron, but to the oil. So far as protection to iron is concerned, no paint con- tains any other protective power than that given to it by the oil, and only by hardening the oil and protect- ing it from the action of the air when it has once be- come sufficiently hardened has any paint protective influence on iron. It is necessary to lay down this prime principle in the strongest way, in order that we may choose our paint with a clear mind, not befogged with notions about affinities or ^'anti-corrosion paints.'* An anti-corrosive paint would be a paint of zinc, for instance, which should have such an attraction for oxygen as to cause the oxygen to attack and destroy it in place of the iron. A zinc bar left in a boiler was found to have entire- ly disappeared, as was supposed, through its protective power over the iron from oxygen. Ruolz found that zinc-covered iron rusted more rapidly than naked iron under all bare spots. Eisner found that tin, iron and zinc together set up The Protection of Iron, 157 such a galvanic action as to crystallize the tin into thin plates which could then be rubbed off. An " antacid'* paint might be used, since iron with- out the aid of acid hardly rusts at all ; but since the amount of carbonic acid in the air is unlimited, we should require an unlimited amount of base substance to neutralize the paint ; and if the surface of the base substance were not in some way constantly renewed, one paint would soon act as it does usually, to wit, as a mere water-proofer for the iron. > At present all paints protect iron by inclosing it in a water-proof layer, continuous and impenetrable. The only question involved in painting iron is the du- rability of this impenetrability. The sun injures the layer in two ways. If, says Prof. Max Pettenkofer, the not-drying oil acids are removed from freshly dried linseed oil by ether and etherial oils, an elastic caoutchouc-like substance remains, which by degrees is hardened and made brit- tle by the air, and in this condition easily separates into parts, or, in simple language, cracks open. Says G. J. Mulder, there are two periods in the drying of linseed oil, the first of which only I have studied. The first period occurs in the early months, and leaves the not-drying part of the oil (20 per cent.) unchanged, only the glycerine ether being driven off and oil-leather (linoxyn) produced. So long as this period lasts is the covering always dry, but remains elastic. When the second period is entered upon, it becomes brittle. The first period lasts a longer time in lower temperature and out of the direct rays of the sun. Under these conditions the oil in drying always increases more in weight. In ordinary temperatures i^S Painting and Painters Materials, and in diffused light loo parts of linseed oil become III to 112 parts, but warmed to 170% 4 to 5 parts are lost. In the direct sunlight 100 parts of the oil gain less than 7 parts ; oil dried in the sun and partly cov ered by the sun gained respectively 5 per cent, and 10 per cent., or a difference of 5 per cent.* Or, in other words, a difference of about one quart of oil in 1,000 ft. of surface. This quart of oil is not- drying oil, which has been changed by the sun's rays into a flying-oil acid during the first three months. Now, supposing the rate f of loss to continue for 12 months, the oil on the 1,000 ft. of surface (20 quarts) would have lost about all of its not-drying acids (4 quarts). Oil drying on the inside of the house would at the same time have lost almost no substance. By actual examination Mulder found that oil in the house con- tained after four months drying its full amount (20 per cent.) of not-dry acids, combined with 91 per cent, of tough oil-leather. Here then are the simple conditions under which we must work in preserving iron by oil : We must prevent the drying part of the oil from becoming hard-dry ; we must keep the soft-keeping not-drying acids from flying away in so great quantity as to re- duce the oil to a brittle mass. To attempt these things is to work on a basis of intelligent facts. Put- ting aside all ideas of affinity "J between this and * The figures are gross, and the calculation of percentage the writer's. Mulder means (he must always be taken as a whole book, not by texts), that the oil acids (not drying) still remain — are not as a body changed. See tables of oil drying. + Of course the rate would decrease, because the hard surface layer would protect the underlying oil. X Except as regards the galvanic action between metals decomposing water. There is also some evidence that iron is protective of iron by coa- nection, but it is, to say the least, capable of two interpretations. The Fi'otection of Iro7i, 159 that paint and iron, the only intelligent questions we need ask about protection by any paint substance are: Does it contain any hurtful acid ? Will it protect the elastic qualities of the oil (the drying and the not- drying acids) from the action of oxygen ? How long will it do this, and what is its price ? Particles of paint protect oil in three ways : ist. They cover the oil and by their own substance thus protect the oil drops from the sun, the air, and the storms of rain, sleet and dust. For this purpose we must, of course, have sub- stances which are not themselves affected by the sun and the atmosphere. Iron, lead, silicon and slate are not much affected, zinc more than other sub- stances when acid is present in the air. Sanding the surface of paint is merely applying the protecting power of silex particles to the surface of the oil, and it furnishes a cheap and efficacious method. Clark tells of a larch pillar in a damp situation, the cracked end partially destroyed at base, but the paint, which was sanded, is still perfect after 35 years. " Sanding," says Stevens, "if well done, will last many years and need not be repainted except to freshen the color or to change it, or for the purpose of cleaning off the stains from dust which may accumulate. The sand should not be the finest nor the coarsest ; well dried and sifted into the third coat, if a new building, with a sander — the best are made like a grocer's scoop, with beveled part of perforated tin, the holes about 1-16 part of an inch in size, and should be made to contain when full about four pounds of sand * * * Machines have been used for sanding to a good advantage, but I prefer the more laborious mode of sanding, as I think the work will be better done by it, besides the machines are very inconvenient. * * The paint in which the sand is to be sifted should be mixed with nearly all oil, and be put on as carefully as if for finishing coats, and the sand must be applied while it is fresh and sifted i6o Painting and Painters Materials. against the surface as long as any appearance of the oil remains. The workman should examine carefully for any greasy spots and dash on the sand again before allowing the paint to dry or set, even. Care must be observed to keep the stretch, or the edge, al- ways beyond or out of the way of the dashing or falling of the sand, for if the brush comes in contact with the sanded part, the work will be unavoidably disfigured or spoiled. Once sanding is seldom sufficient for a good solid look of the work ; a second sand- ing, the paint should be a Httle thinner than for the first. These directions apply as well for wood as for brick work." — Stevens Art of House Fainting. In general, however, painters have preferred to mix the protecting substance with the paint. The protection of such mixed substance is, however, greater, if the paint particles are very fine, and it is more perfect the more perfectly the particles are com- bined uniformly with the particles of oil. For these reasons iron oxide paints should first be ground very fine, and then thoroughly ground for hours with the oil. Care should also be taken that they contain no acid (as many iron ores do), and that the water has been thoroughly roasted out of them. 2d. There is a chemical combination between the oil and paint as soap, which is protective. Lead cer- tainly (except carbonic acid lead), and probably iron and zinc, to a degree, unite with the oil, and hold fast the flying acids. It is very difficult to bring proof on every point, but it is probable that where no portion of the protecting substance unites with the oil, there is greater tendency of the paint to crumble. At least it is true that all substances used as pure paints have some soap-making powers, and wherever hardness is required soap must be had, or the hardness Painting Iron, i6i will be the over-drying of the oil, or the first step in decay. The hardest soap is given by lead, especially red lead, and (3d) this hardness is a protection by itself. When a freshly painted car or carriage is placed in the sun it is injured, while the same vehicle may be freely exposed after two or three weeks' hardening. The harder the immediate surface, the less effect pro- duced by the air and the sun on the oil underlying this hard layer. We come now to the final question in regard to paint for iron, namely, price. I think it may be safely said there is no paint equal to red lead for wrought iron 7iot thoroughly clean of all kinds of rust. For per- fectly clean iron and for cast iron, iron oxide paint of good quality may be quite as good. This is quite certainly the cheapest paint. This chapter is written to show the elements of a good paint, not to advocate the use of red lead. Properly hardened iron paints, all considerations in view, should be among the very best for iron and wood — the violet-colored for iron^ the bright red for wood, (See Pigments, Part II.) Exceptionally, the more expen- sive and injurious red lead may be used alone. Iron and white lead make a like hard mastic (Seddens). The Oil. — The value of a paint is its oil, and econ- omy consists in purchasing good oil-the best and purest to be obtained. Adulteration will be advantageous in some cases, but it should be done for a purpose. On roofs 15 to 20 per cent, of fish oil will be an advan- tage, and on bridges which are not so frequently re - painted,probably a little cotton-seed oil will do no injury. j62 Painting and Painters Materials The principle of adulteration is very simple : the more " elastic " the first coat the more durable, but the greater the tendency to crack the second coat. The weather coat should always be the more " elastic ' (containing more not-drying oil), but a man of judg- ment will discriminate. The most elastic (in the true sense) of all oils is boiled oil, but it is too thick, and needs to be thinned with raw oil or with turpentine. A priming of such oil thoroughly dried will stick, if red lead be added, and if not put on too thick. A little boiled oil thoroughly mixed with raw oil in- creases its elasticity, but probably increases the chances that the paint will wrinkle, pit and crawl a little, al- though if care be taken in the mixing and using, any such effect should be so small as to be imperceptible There is this strong reason for using more or less boiled oil, that the pigment — at least lead-pigment,will more thoroughly unite with and harden the oil, as boiled oil is partly composed of free linseed-oil acid, and we shall thus get more hard soap. At the same time we thus get a harder oil ; we always get with the boiled a certain quantity of oil rubber produced by the boiling, and which so long as it remains gives the oil a rubber-like elasticity. It must never be forgotten, however, that both the hardness and elasticity are pur- chased at the price of color, and that such a paint will certainly yellow — a fact of little consequence, how- ever, unless white or delicate tints are to be used. The evidence in favor of boiled oil is very clear in many respects. We shall take up the matter in another chapter. CHAPTER XVI. Brushes. The qualities required in a brush are elasticity com- bined with firmness and softness. Elasticity is the prime requisite of all brushes ; brushes for finer work require a great degree of softness. No available substances but the bristles and finer bristles (hair) of animals have these qualities. Hogs Bristles. — Those of the wild boar of Russia are the best for two reasons, (i) The hair of North- ern animals is longer than those of warm regions. (2) But the Russian bristles are superior mainly be- cause the animals from which they are taken are much older. Our pigs are allowed to feel only one or two winter's cold before they become pork. The Russian pig, however, may see ten or fifteen or twenty winters, and in general the older he is the better his crop of bristles ; they are longer and stiffer, the best Russian bristle being the Okata," which is very stiff, and six or seven inches in length. The next quality, or Firsts, is of medium stiffness and four to six inches in length. The third quality, Suchoy, is from 4^ to 5^ in. in length, the fourth quality, " Seconds," being one inch shorter. The last two grades are known as soft bristles. The finest of short bristles come from France, a very fine quality from the South of France, of which there is only a limited supply. White bristles are much more valuable in the mar- ket, bringing one-third more in price. They have ap- 164 Painting and Painters' Materials, parently no advantage over dark bristles except the greater difficulty of adulterating them. The white bristle has not only to be sorted out from the darker ones but they also require bleaching. Many bristles, both white and black, are taken from the backs of Western hogs. These, for reasons given above, are a very inferior article. The importation of foreign bristles is about 1,000,000 of lbs. They come from Russia, Poland, Belgium and France, and are all superior to our home production — good pork and good bristles are not found together. In Russia, the bristles come from two sources: (i) Slaughtered animals. (2) From the woods and wilds, where they are found and gathered by the poorer and more shiftless class of the people. The boar sheds his bristles at least every year, if not oftener, rubbing them off at his favorite scratching posts. Here they are found, bunched and sold to dealers, who take them to factories where they are sorted and cleared from decaying flesh, etc. The brush manufacturer bleaches (with brimstone, etc.), combs, and sorts the bristles as to length. Adulteration. — Bristles are adulterated with inferior American bristles and with Tampico," a kind of grass, imported from Central America. This makes a harsh, brittle brush and is in every way an inferior substance to any kind of bristle. A grass from Florida is also used for shorter bristles. Tampico is of a dark gray or blackish color ; the Florida grass white. Bristle Brushes. — A prime quality of a bristle is its flag" — the division of the bristle on the end into Brushes. two or more fingers." The best older Russian bristles have several fingers in many of the " flags." The flag adds fineness to the surface, while the coarser body of the bristle gives elasticity and stiff- ness to the brush. A brush-maker of long experience tells me that by using always the same sides, so that the brush is worn chisel-shape (not stubbed to a point), the flag will split down the shank of the bristle as the brush wears away, and the painter will get the benefit of it for a longer time. At least, a brush should be worn oval or to a flat chisel point, never to a sharp point, such as is produced by constantly turning the brush in the hand. A stubb " is a sign that a green hand is at the other end. There are those who object to very long and stiff bristles, because after being somewhat worn they throw and spatter the paint. A brush of medium length is the best. Brushes are numbered i-o, 2-0, 3-0, etc. No. 6-0 is a proper size for outside (house) painting, and it is well to bridle it about one-third down with twine, unwind- ing as the brush wears away.* Inside priming is best done with the same brush, well worn. Finer Brushes. — The finest, softest and most elastic brush is made from the hair of the sable. A genuine sable brush, however, is very expensive, a small short brush (they are usually short) being worth about four dollars. Fitch Brushes, — The next grade of brush is, per- haps, the best, genuine fitch. The fitch is a small an- imal like the weasel. There is but a small supply of such brushes in the market. * The brush should not be bridled too tight, or the paint will squeeze out and smear the hand. i66 Paintiiijj^ and Painters' Materials, Skunk Hair Brushes. — This unpleasant little animal supplies much hair for imitation fitch and even sable brushes. Badger hair brushes are longer and stiffer. Bear hair brushes are also used. CameVs Hair Brushes. — These are made entirely from squirrel's hair. They are soft and fine, but lack elasticity. The Use of Brushes, — This is so entirely a matter of **tact" that each man will find out his own best way. Not so always with the choice of brushes. Copper-bound brushes are the best. Good brushes are always the economical ones. The Care of Brushes, — Dry brushes should never be kept in a warm or dry place, as they are liable to shrink. Paint and varnish brushes should not be used dry, and never as dusters. Paint brushes should be hung {^without touching bottom) in water, from a wire or other frame, in order that the water may reach some distance toward (not above) the bridle. The bottom of the tub will soil the brush, for it is quite certainly dirty. Varnish Brushes. — These should be kept in oil and turpentine or in varnish. Brushes used for English varnish must be kept in varnish. This will save a deal of mystery." Varnish brushes should be thoroughly cleaned from time to time in order to prevent their becoming ^Mousey" — /.