See other formats

Full text of "Rustless coatings; corrosion and electrolysis of iron and steel"

Google 



This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project 

to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 

publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing tliis resource, we liave taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 
We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for in forming people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement liabili^ can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 

at |http: //books .google .com/I 



I. 

i 



UBRARv 



Iron Column nkau Kotab Misoii. DELHr, India. ERECTXii b.c. 90(1 
Protttupieee. 



RUSTLESS COATINGS; 



CORROSION AND ELECTROLYSIS 
OF IRON AND STEEL. 






Member of the American Society of MechanUxU Engineers, American Association 

for the Advancement of Science; Consulting and Otntracting 

Engineer; Superintendent of Motive Power, late 

United States Military Raihvays. 



FIRST EDITION. 
FIRST THOUSAND. 



NEW YORK : 

JOHN WILEY & SONS. 
London: CHAPMAN & HALL, Limited. 

1904. 



Copyright, 1904, 

BY 

M. P. WOOD. 



• • 






I • * • 
• • •• 



BOBKBl' DBUmiOND, PHUtTlER, MCW TOIUL 



t 






(i 






/2^' 






PREFACE. 



Since the publication of the papers ''Rustless Coatings for 
Iron and Steel/' in the years 1894-1901, the author has received 
many requests from engineers and others to present them in a 
more available form than in the Transaclions of ike American 
Society of Mechanical Engineers and in the columns of the various 
American and foreign technical journals. The subjects have been 
mainly rewritten and new matter added to bring them up to date. 
The characteristics of oils, pigments, and paints that form the prin- 
cipal protective coatings for ferric and other structures are given 
dt length under their respective chapters. It is believed by the 
author that the collected data will afford a reliable source of in- 
formation of what paints are composed and what may be expected 
of them. The technical journals have given a great deal of space 
to the subject of proctective coatings for metals, but the data are 
not always available when comparison with some recent result or 
experiment is required. 

The author acknowledges the assistance afforded him in the collec- 
tion of the data, by the technical press and nearly two hundred other 
sources, and he has, as far as possible, given credit for the same. 

The subjects are so grouped in the work, and so detailed in the 
index, that the busy man can find the data that will bear on the 
case in hand and enable him to avoid some, if not all, of the un- 
favorable results that have attended previous applications of some 
misleading trade-mark mixture. Most of the analyses and tests 
of the commercial pigments and paints have been repeated many 
times without any material discrepancy from the data herein given. 
The results from the use of many of these paints are apparent in the 
excessive and continual corrosion of important ferric structures 
everywhere. He that runs may not always read, but he can at least 
see; hence should be able to proJ5t by the experience of those who 
have preceded him. 

M. P. Wood. 

• • • 

111 



CONTENTS. 



PAOB 

CHAPTER I. 

Paints, of What Composed. How Destroyed. Classified as True Pigments 

and Inert Substances 1 

CHAPTER II. 
Paint Statistics and General Character 17 

CHAPTER III. 
Iron. Iron Oxide. Copperas. Ochre. Umber. Spanish Brown 26 

CHAPTER IV. 
Red Lead. Litliarge 45 

CHAPTER V. 
White Lead. " Old Dutch," Quick Process, Electrolytic, and Sublimed Lead. 61 

« 

CHAPTER VI. 
What Constitutes a Good White Lead 74 

CHAPTER VII 
Zinc Oxide. Sulphate of Zinc. Lithopone and other Zinc Pigments 89 

CHAPTER Vin. 
The Carbon Group of Pigments Lampblack 98 

CHAPTER IX. 
Mineral and Artificial Asphalt. Coal-tar 103 

V 



VI CONTENTS, 

CHAPTER X. 

PACK 

Asphaltum Paints and Carbon Varnishes. Fossil Resins 110 

CHAPTER XI. 
Baked Japan Coatings lig 

CHAPTER XII. 

Dr. Angus Smith's Anti-corrosive Water pipe Coating and other Water- 
pipe Dips and Coatings 123 

CHAPTER XIII. 
Graphite. Amorphous, Flake, and Acheson's Electric Furnace 13ft 

CHAPTER XIV. 
Bessemer Paint and Furnace-slag Pigments 145 

CHAPTER XV. 
Hydraulic Cement Coatings and Concrete 149 

CHAPTER XVI 
Bower-Barflf Coatings 166 

CHAPTER XVII. 
Qalvanlzing Processes 172 

CHAPTER XVIII. 
Inert Pigments and Adulterating Substances 184 

CHAPTER XIX. 
Spirits of Turpentine 193 

CHAPTER XX. 
Bisulphide of Carbon. Tetrachloride of Carbon 203 

CHAPTER XXI. 
Japan Driers , 208 



CONTENTS. vii 

CHAPTER XXII. 

PAGE 

Flax-plant. Linseed and Linsecd-oil 21 1 

CHAPTER XXIII. 
Boiling Linsecd-oil. Processes and Experiments 225 

CHAPTER XXIV. 
Drying of Linseed -oil and Paint Coatings 286 

CHAPTER XXV. 
Tests for Linseed-oil and Adulterants 240 

CHAPTER XXVL 

Substitutes for Linseed-oil. Lucol. Euphorbium. Chinese Wood-oil. 
Japanese and Chinese Lacquers 248 

CHAPTER XXVII. 

Decay of Paint. Catalytic Action of Pigments. Caustic Action of Mortar 
and Cement 261 

CHAPTER XXVIII. 
Sand-blast and Pickling Processes for Cleaning Mctiil 271 

CHAPTER XXIX. 
Paint Tests. Toltz's. Smith's. Sulvay Cos*, and others 279 

CHAPTER XXX. 

Tests of Paints on Ferric Structures Spennrath's Experiments on Paint- 
skins 295 

CHAPTER XXXI. 
Painting by Spray 304 

CHAPTER XXXII 
Mixed Paints. Enamel Paints and Baking Enamels 310 

CHAPTER XXXIII. 
Corrosion of Iron and Steel 322 



vm CONTENTS, 

CHAPTER XXXIV. 

PAGB 

Electrolysis of Underground and other Ferric Substances 370 

CHAPTER XXXV. 
Marine, Anti-corrosive, and Anti-fouling Paints and Ferric Alloys 400 

CHAPTER XXXVI. 

Table of Pigments and Inert Substances 408 

Characteristics of Metallic Bases of Pigments 410 

Gases and Elements that Cause the Decay of Paint 411 

Oxygen in Pigments. Combinations of Oxygen, Carbon, and Sulphur 413 

Changes in Pigments Due to Atmospheric Influences 413 

Corrosive Elements in Snow-water, Smoke, and Oils. Saturated Air. 414, 415, 416 

Menstruums. Oils and Solvents '417 

Fatty Acids and Solvents. 418 

Proportion of Oil in Pigment Pastes 418, 419 

Electro-chemical Elements 420 



LIST OF ILLUSTRATIONS. 



Frontispiece : Iron Column of Dellii. (Page 170.) 

no. PAGB 

1. Covering Power of Paints 4 

2. •* ** " " 5 

3. Cellular Formation of Wood (Peeling of Paint) 12 

4. Corrosion of Iron. Atmospheric Exposure 28 

6. •' " ** Single Application of Water 29 

6. «4 i* Tin-plate in a Damp Cellar 38 

7. Mill-scale Corrosion 40 

8. Film of Red-lead Paint incrusted with Rust 51 

9. ** " ** ** porous with Air-bubbles 53 

10. *' " " ** dried on a Glass Plate 55 

11. Red lead, White Lead, and Zinc Oxide Paint-skin 56 

12. White Lead. Corroding Pot and Lead Buckle 62 

13. *< ** Carter's Process for Corroding 68 

14. '* " Action of Sulphurous Gases on White Lead 77 

15 and 16. White Lead and Sublimed Lead (>)atings on Picket Fence con- 
trasted 86 

17 and 18. White Lead and Sublimed Lead Painted Boards 87 

19. Fossil Resin. Section of the Resin Ill 

20. " " (Lithograph) Ill 

21 . Pipe dipping Tank 1 23 

22. Boiler-tube coated with Graphite Paint 141 

23. Acheson's Electric Furnace 143 

24. Efflorescence on Brick Walls 163 

25 and 26. Hot and Cold Process Zinc Coating (Galvanizing) 173 

27. Table of Zincing Solutions *: 176 

28. " ** One-minute Immersions for Test of Coatings 179 

29. Inert Pigments- Covering Power 187 

30. Boxing Trees, for Turpentine, Old M» thod 195 

31. " '* *' " New Method ..201 

32. Flax-plant Blossom from Jerusalem 212 

33. ** Flower, Seed-vessel, Seed and Root 214 

34. Oil-seeds 221 

36. Mill-scale Corrosion, Railway Viaduct 262 

36. ** ** Phoenix Column 262 

37. Corrosion of Steel Girder, Washington Street Bridge 263 

88. ** ** Sidewalk T Beams 269 

ix 



X LIST OF ILLUSTRATIONS. 

no- PAGE 

39. Sand-blast Apparatus. 271 

40. Portable Sand-blast Macbine 272 

41. Field Spray Apparatus at Work 305 

42. Helmet for Spray Painting or Sand-blasting 306 

43. Barrel and Hand power Spray Apparatus .307 

44. Hand Spray Apparatus . 308 

45. Power Paint-mixer 311 

46. Pipe Coated with Sublimed Lead 317 

47. Corrosion of a Laminated Steel Girder 333 

48. ** " Plate Girder 335 

49. •* •* Tee Rail in a Tunnel 336 

50. '* ** *• ** on a Dock 338 

51. »* Increased by Stress 349 

52. Diagram of the Corrosibility of Metals 354 

53 and 54. Diagram of the Stress Corrosibility of Wrought Iron 356 

55. Diagram of the Stress Corrosibility of Cast Iron 357 

56. •• " " *' •• »• Cast Iron in Compression 358 

57. • • * * ' * " • * * * Hard-draw^n Copper Wire 358 

58 Corrosion of Mi'd Steel (Ammonia Chloride Solution) 364 

59. •* *• Cast Iron (Polished) 364 

CO. •• " •* " (with Scale) 365 

61. •• •* Mild Steel (Ammonia Chloride Solution) 365 

62. *• '• Burned and Hardened Steel 365 

03. • • " Burned Steel not Hardened (with Scale) 366 

64. •* •* Annealed Steel (Polished) 366 

65. •* *• *• " (with Scale) 366 

66. ** * * Steel Burned but not Hardened (Polished) 367 

67. '* " Hardened Steel (Polished) 367 

68. ** *' *• " (with Scale) 367 

69. *' " Steel Burned and Hardened (Polished) 368 

70. " " Sheet Iron (Polished) 368 

71. ** *» *• *• 368 

72. " ** " • ' (with Scale) 369 

73. * ' ' * Peoria Water Stand-pipe. Pitting of Sheets 374 

74. *' " " " •• Course of Electric Currents. . . 376 

75. " " " " *• Interior View of Inlet Pipes. . 376 

76. * '* a 12-inch Cast-iron Water-pipe 377 

77. Electrolysis of a 4-inch Cast-iron Pipe 378 

78. " '• Cast-iron Water-pipes, Reading, Pa 380 

79. " *' 16-inch Suction and Water Mains 381 

80. " *' a 6-inch Water-pipe. Providence, R. 1 382 

81. ** •♦ End Posts of a Swing-bridge Truss, Providence, R. I.. . . 383 

82. •' "an 8-inch Water-pipe, Springfield. Ill 384 

83. " ** Lead Service-pipes and Telephone Cable Coverings, 
Brooklyn. N. Y. 385 

84. Electrolysis of 6-inch Water-pipes, Kansas City, Mo 386 

85. '• " Street Railway T Rail, New York 386 



RUSTLESS COATINGS. 



CHAPTER I. 
paints: of what composed, how destroyed, classification as 

TRUE pigments AND INERT SUBSTANCES, ADULTERANTS, ETC.* 

What is paint? This question can be answered in a broad way 
by saying: It is any liquid or semi-liquid substance applied to any 
metallic, wooden, or other surface, to protect it from corrosion or 
decay, or to give color or gloss, or all of these qualities, to it. 

A better definition would be, that paint is a compound of a pig* 
ment and a liquid, usually appUed to any surface with a brush, for 
the purpose of protection, or to secure artistic effects; which liquid, 
after undergoing certain changes, in part mechanical, or chemical, 
or both, has the power of holding the pigment to the coated surface. 
It is evident that the latter definition would also include those com- 
pounds which are applied to many surfaces either hot or cold as a bath, 
rather than by a brush, solely as a matter of convenience or rapidity; 
and particularly so when metallic members of large size, or with 
intricate and hidden parts, are to be protected. 

The essentials of a good paint, for whatever use intended, are: 

First, — ^That it shall adhere firmly to the surface over which 
it is spread, and not chip or peel oflf. It must be non-corrosive to 
the material it is used to protect, as well as to itself under long periods 
of atmospheric exposure and chemical changes. It must form a 
surface hard enough to resist frictional influences, yet elastic enough 
to conform to all changes of temperature, or with a coefficient of 

* Excerpts from a paper by the author presented at the Detroit meeting 
(June, 1895) of the American Society of Mechanical Engineers, and forming 
part of Volume XVI, paper number 637, of the Transactiona, 



2 ESSENTIALS OF A GOOD PAINT. 

elasticity approximately as near the material it covers as possible. 
It must be impervious to and unaffected by moisture, atmospheric 
or other influences to which the structure may be exposed. 

Second. — ^That it shall work properly during its application — a, 
property which depends largely upon the relative amounts of pig- 
ment and liquid. The natures of both pigment and liquid also have 
influences that govern results. 

Third. — ^That it shall dry with sufficient rapidity. This func- 
tion depends mostly upon the vehicle or liquid used with the pig- 
ment, though the pigment has in many cases an influence, as will 
be seen further on. 

Fourth. — That it shall have proper durability, which is a func- 
tion both of the pigment and liquid. And as the question of cost 
is in many cases the governing factor in the selection of a paint, 
the question of durability may be regarded as the most important 
one of the list. It should be understood, however, that a paint can 
be durable per se, and not be protective in the strict sense of the 
word, as can be illustrated in the case of a good paint applied to the 
surface of a sheet of iron coated with rust. The Uquid element in 
the paint ^dll not absorb or neutralize the corrosion which it covers, 
but will dry regardless of it, and permit the destruction of the metal 
to progress beneath its coat. 

Fifth. — Covering power, by which is meant the power of a pig- 
ment so to cover the surface to which it may be applied that its pro- 
tection from decay is not only assured, but that the minimum amount 
of paint shall effect this purpose. 

Th£ covering power is also used to express the power of a pigment 
to protect the oil from decay, in which case a large amount of pigment 
and a small amount of oil are used. This description of paint dries 
more or less "flat," the pigment being exposed to the weather and 
held in place by the thin film of oil. It is thought by many master 
painters that this is the most durable and best paint for general use. 
On the contrary, paints that dry with a gloss have a large amount 
of oil and a small amount of pigment, in which case the oil covers 
and protects the pigment. 

It may.be used to express the amount of color upon the surface; 
as, generally, if a surface has plenty of color upon it the covering 
power is said to be good. To illustrate this definition: If an iron- 
oxide paint is proportioned so that the ratio between the pigment 
and the oil is by weight 50 per cent of pigment and 50 per cent of oil 



PAINTS, COVERING AND COLORING POWER. 3 

when the paint is ready for spreading, and the pigment consists of 
30 to 40 per cent of iron oxide, the covering power will be said to be 
good; but if the same proportions of 50 per cent ratio between the 
pigment and the oil be had, in which the iron oxide is only 5 per cent 
of the pigment, the covering power would be called poor; and so it 
would be in the case where 10 per cent of pigment and 90 per cent 
of oil were used. If in the two latter cases the oil contained large or 
liberal amounts of volatile diluents, the appearance of the surface 
would indicate a deficiency in the covering power of the paint. 

The covering power is also commonly expressed in the amount 
of surface which a given weight of paint will cover. A good iron- 
oxide paint will cover nearly twice as much surface as -white or red 
lead. The specific gravity of the paint also is to be considered in 
the definition of this power. The lightest paints haVe the most 
covering power. White lead is about 6.4 times as heaVy as water; 
iron oxide 5 times; yellow ochre 3^ to 4 times, etc., ete. With this 
variation it is manifestly almost an impossibility to get the same 
number of particles of the same size out of the same weight of different 
materials. 

Fig. 1 represents the covering power of a number of paints and 
inert pigments. 

The refracting power of light has much to do with an under- 
standing of this covering power of paint. The greater the refracting 
power of the pigment is over* that of the oil, the better will be the 
covering power. The index of refraction of air is 1 degree; water, 
1.34; linseed-oil, 1.48; glass, 1.50 to 1.55; silica, 1.55; feldspar, 1.54; 
whiting, 1.65; chrome-yellow, 3.00; vermilion, 3.20, etc. There 
is no exception to the rule that the finer the state of division to which 
any pigment is reduced, the better will be its covering power. Sul- 
phate of lime, barytes, feldspar, silica, talc, whiting, etc., are all of 
low refractive power, and of themselves, independent of this refrac- 
tive quality, do not constitute good pigments, though when mixetl 
with the metallic pigments and ground together in the oil the result 
is a pigment of good covering power, almost as good as the best one 
of the combination. For instance, 80 per cent of sulphate of lime 
and 20 per cent of zinc white, form a pigment almost as good as all 
zinc white, and 10 per cent of whit^ lead and 90 per cent of talc care- 
fuDy groimd, give a very satisfactory result so far as relates to the 
covering power ; but all of the above and other kindred compasitions, 
while improving the covering pow'er, are to be classed as adidterants, 



4 PAINTS, REFRACTING OR COLORING POWER. 

the use of which is objectionable bo far as durability and protective 
power are concerned. 

The covering power is due to two qualities. For instance lime 
whitewash has very little covering power until it becomes dry. 
Barytes covers well as a water paint, because the water leaves it as 
a dry powder on the coated surface. But barjrtes covers poorly in 



Fig. 1. — Covering power of paints. 
oil, because the oil remains with it, and the light reaches it through 
the transparent film of oil. 

Prof. Von Bezoltl's experiments,* from which I quote, illustrate 
how lime, barj-ies, white lead, and other crystalline pigments, when 
mixed in oil, become more or lees translucent, and therefore do not 
color the surface that they cover. 

*Voa Bezold. "Hieory of Color." L. Prang Co. Gondict'e "PMDttngand 
Painting Materials." — The Railroad Gazette. 



PAINTS, REFRACTING OR COLORING POWER. 5 

"If a Bmall glass test-tube be partly filled with powdered glass, 
the powder will appear white, and it will be impossible to see through 
it, but as soon as water is poured into the tube, the powder, to a 
certain d^ree, becomes translucent. By substituting turpentine for 
the water, the degree of translucency is materially increased. If a 
email quantity, of sulphuret of carbon is added to the turpentine, a 
liquid is obtained that reflects (bends) the light about as powerfully 
as glass. If some of this liquid be poured upon the powder in the 
test-tube, the powder will disappear, and light passes through the tube 
as freely as though no powder were present. 

If a solid glass rod be immersed in such a liquor (or a mixture of 
olive-oil and oil of cassia), it will appear as if the rod only reached to 
the surface of the hquid. Within 
the liquid the rod cannot be seen 
(Fig. 2). 

It is shown by these experi- 
ments that the presence of one 
transparent body within another 
is only detected by the eye when 
the two differ in their power of 
refracting light. 

Many white substances are white 
because they are in fine particles. 
A white lily is white because it 
consists of little cells which reflect 
all kinds of light, again and again, 
until it reaches the eye from some 
part of its surface. Water be- pjg. 2. 

comes white when it is broken into 

fine drops, as in a waterfall or on the crest of waves. White lead 
and zinc owe their whiteness to their dense, fine, powder-like condi- 
tion, and transparent glass becomes white when finely pulverized," 

As stated before, the finer the pigment is subdivided, whether as 
a paste which is afterward thinned with oil or volatiles to a consist- 
ency to spread with a brush, or is ground in the oil direct (a procesa 
that all pigments will not endure without injury to their color — the 
scarlet lead chromate, for instance) to the proper consistency to 
spread, the better will be its covering power. 

An ounce of lampblack, because of the minutenesB of its par- 
licles, will cover more surface in an effectively protective maimer 



6 PAINTS, REFRACTING OR COLORING POWER. 

than any known pigment, and one part lampblack and nine parts 
sulphate of lime by weight, give most excellent results in covering 
power. Prussian blue, the scariets, lak^, and others of what can 
be called "the fugitive colors," on account of their tendency to fade 
out, possess the lightrdispersing power which deceives the eye as to 
their covering power, when in reaUty for actual covering as protective 
substances they are absolutely worthless. These colors should be de- 
nominated stains rather than paints; for generally the only measure 
of protection from decay or corrosion which accompanies their use 
IS solely from the oil or liquid with which the color is mixed. 

The designing of a paint, for whatever purpose, necessarily in- 
cludes the qualities already mentioned, viz.: adhesion and elas- 
ticity, working qualities, drjdng qualities, durability, covering power. 
The other quality, the cost, cannot be ignored, and will be duly con- 
sidered later, as well as what pigments to use for the intended purpose. 
All pigments do not contain all of the above qualities. The question 
naturally arises: Is it necessary for a pigment to be pure and unmixed 
with inert substances, or can a certain amount of these be mixed 
with the pigment without detriment to it? 

Experiments of long duration lead to the conclusion that the 
oxides of iron, lead, manganese, and other strong pigments, can be 
mixed with large amounts of these inert substances without detriment, 
and generally to the manifest improvement of the paint as a pro- 
tective agent on many structures, notably wooden or composite ones, 
A single illustration will suffice to make this apparent. Oxide of 
iron is one of the strongest of pigments in covering power. If one 
ounce of this pigment be spread in two coats over a given surface, 
say two square feet, so that the surface be completely hidden, and 
the job be declared a satisfactory one so far as covering power is con- 
cerned, and in the second case an ounce of the same oxide of iron be 
mixed with three ounces of barjrtes, kaolin, gypsum, etc., and this 
paint be spread over two square feet of surface as before, it is obvious 
that the amoimt of color per unit of surface will be the same in both 
cases; but in one case there is four times as much pigment as in the 
other, and in the second case three-fourths of the paint would be 
inert material. For railway cars and wooden structures the dura- 
bility of these paints would be in favor of the second case, as well as 
the cost of the paint. The pigment in this case is the life of the paint, 
and protects the oil from the decay incident to oxidation from atmos- 
pheric exposure. 



PAINTS, COVERING AND COLORING POWER. 7 

Oxide of iron is practically unchanged after centuries of exposure. 
It induces and promotes oxidation in all organic substances with 
which it is brought into contact, and in nearly all metallic bodies. 
In an oxide-of-iron paint it is the oil which decomposes, it being 
the organic matter. The decomposition is due to the exposure to 
the elements aided by the oxidizing power of the oxide of iron pig- 
ment mixed with the oil. This statement holds true only where 
there has been no chemical change or combination between the 
pigment and the liquid. 

Whiting, sulphate of lime, barytes, kaolin, silica, feldspar, and 
talc are the principal inert substances used in pigments. Whiting, 
gypsum, and barytes are the best of the list; the others, grinding 
greasy, are hard to grind, or of a nature readily decomposed by 
water, are objectionable. Barytes, from its great weight, is objec- 
tionable as a paste or prepared paint. Its use as an adulterant is 
given in Chapters VI and XVIII. The sulphate of lime (gypsimi) is 
no doubt the best of the inert substances to mix with any pigment, 
all things considered. It should be thoroughly hydrated. As high 
as 45 per cent by weight of this substance can be mixed with 50 per 
cent of sesquioxide of iron for a pigment. Many of the oxide-of-iron 
paints are made by ignition of copperas, and a notable amount of 
sulphuric acid is usually left in the oxide which the heat has failed to 
drive off. From 2 to 5 per cent of carbonate of lime is added to neutral- 
ize the free acid, changing it to sulphate of lime. In these proportions, 
the pigment really consists of 50 per cent of oxide of iron and 50 per 
cent of inert material, all by weight. Any oxide-of-iron paint which 
contains hydrated oxide or free SO, will deteriorate rapidly by oxidiz- 
ing the liquids, while any free SOj will retard the dr3dng of the paint. 

A good paint prepared for spreading in ordinary temperatures 
upon wooden or composite structures has the ratio by volume of 
about one-third pigment and two-thirds oil or liquid. The practice 
upon one of the leading railwa3rs of the United States, where the 
materials purchased for paints amount to over $300,000 yearly, is 
to allow 75 per cent of pigment and 25 per cent of oil by weight, for 
the paints applied to cars and wooden structures. 

Experiments determine that the most durable paints are those 
which contain a large amount of pigment per unit of surface; and that 
pigment is the best which is strong enough of itself, or with a proper 
proportion of inert material, to allow liquid enough to be added to 
it to flow and work well with the brush when applied. 



8 PAINTS, MECHANICAL INJURY. 

Destruction of Paint. 

The destruction of paint may be from eight causes: First, mechan- 
ical injury; second, the action of deleterious gases; third, chemical 
action between the pigment and the vehicle or liquid; fourth, chemical 
action between the body covered and the paint, either the pigment 
or the liquid; fifth, the action of light; sixth, peeling; seventh, destruc- 
tion by cleaning; eighth, water. 

Many master painters and manufacturers claim that the destruc- 
tion caused by cleaning and the action of water are the worst of the 
above causes. This is true so far as paint applied to wooden structures 
is concerned, but has no relation to the causes which effect the destruc- 
tion of paint applied to iron or steel structures. As most of the 
above destructive agents are common to all structures (wooden, 
metallic, or composite) which depend in a greater or less degree for 
their preservation from decay or corrosion, upon paint (under which 
name all paints, oils, varnishes, japans, and surfacers are classed), it 
may not be amiss to discuss briefly each of these causes in detail before 
citing the destructive agencies which relate sokly to the corrosion 
of metallic structures, the prevention of which will require the con- 
sideration of other preservative methods than paints, or which may 
be used in connection with paint to secure the best protective results. 

First. — Mechanical injury to wooden structures is not a serious 
cause of deterioration of paint. Near the seashore the wind and sand 
have the effect of a sand-blast, which cuts away the paint rapidly, and 
in this case the more elastic the paint, the less will be the mechan- 
ical injury. This sand-blast action is quite as effective on ferric 
structures, and as generally they are of a more important character 
than the wooden cottages and minor buildings on the seacoast, its 
action must be guarded against. If the paint coating is of a soft, 
spongy nature it will resist the sand-blast, but will absorb moisture 
from the air, and hasten either the oxidation of the paint or the 
metallic surface which it covers. 

A further injury to metallic structures can be classed under the 
head of mechanical, viz.: that arising from the expansion and con- 
traction of the various parts from the atmospheric changes which 
are constantly going on, changes ranging from 40 degrees F. to 150 
degrees F. not being unusual. It is an impossibility to proportion a 
paint compound so that its coefficient of elasticity wdll be the same at 
all temperatures as that of the metal it covers. It may be possible 



PAINTS, MECHANICAL INJURY. 9 

to do this at some temperature at or between 60 degrees and 90 degrees 
F., or even between +40 degrees F. and 90 degrees F.; but that any 
paint m the class of conamercial colors will do this at all temperatures 
is the tale of the salesman. It may be argued that; these changes 
coming from the external surface of the paint and being transmitted 
through its coating, it will be the first to adjust itself to the new or 
varying relation between the metal and the paint, and so will work to 
the advantage of the paint in making the change, this being in ordi- 
nary cases a gradual one. If the paint is of an elastic, close-clinging 
material, and not a hard, vitreous one, the claim will hold good. 

The compounds which most closely partake of this nature, will 
be spoken of hereafter. An addition to this problem will be had 
when the strains due to the action of wind, the passage of railway 
trains, and those due to changes of a sudden and vibratory character, 
together with the action of snow, hail, and water driven at high 
velocities, are added to the temperature changes. These strains 
necessarily come to the metal first, and whatever changes occur 
in the bars by the strain, the paint must accompany them. As 
these strains are generally of a vibratory or percussive character, 
it can easily be seen why they should be classed in the list of 
mechanical injuries. In fact, they are a succession of blows which 
the structure must absorb, withstand, and extinguish within itself or 
its connections; the structure then returning to its normal condition, 
the paint or other protective covering must accompany it, instead 
of loitering by the way and being grounded or "left" in the chain of 
operations. 

Second. — The action of deleterioua gases is very familiar to those 
who have studied paints and protective compounds. Sulphuretted 
hydrogen is one of the most common and active of these gases, and 
is formed in excessive amoimts wherever coal is distilled for illuminating 
gas. Sulphurous acid fumes also, being disengaged in the combus- 
tion of coal in the many arts, transportation, and manufacturing 
processes of the day; gases engendered in workshops, being of a 
compound character carrying anunonia, carbonic acid, nitric acid, 
and other fumes, are active agents of corrosion to metallic bodies, 
also. to the paint compounds that cover them. (See Analysis of 
Smoke, Chapter XXXVI.) 

Third, — Chemical auction between the pigment and the vehicle. 
This is an exceedingly important field of inquiry, and largely an 
unknown one. Many of the siccative and other oils which are in 



10 PAINTS, CHEMICAL ACTION ON. 

common use for paints are capable of saponification. It is well 
known that soda and potash are not the only substances which com- 
bine with fats to produce soap, and that almost any of the bases can be 
combined with the fatty acids of nearly all oils to make soap, hence 
we have iron soap, lead soap, zinc soap, manganese soap, etc. Many 
pigments are simply oxides or hydrates, in the same way that soda 
and potash are, and it is strongly suspected that they combine with 
the oil to form soaps, in which case it will be evident that, after the 
paint has been left on the surface for a niunber of years, instead of a 
pigment held to the surface by the Uquid and which has undergone 
certain changes called "drying," it is in reality a new chemical body 
consisting of the constituents of the liquid combined with the pig- 
ment, or in other words, it may be a soap. 

Fourth, — Chemical action between the body covered and the 
paint, either the pigment or the vehicle. The chemical changes 
which may or do take place between the pigment and the liquid 
as set forth in Article III, can be supplemented here to embrace 
those paints which contain pigments, one or more of wliich give 
up oxygen or break down in the presence of organic matter — the oil 
or liquid of the paint. Hydrated oxide of iron (iron-rust) oxidizes 
organic matter (the oil) and gradually destroys it. Oxide-of-iron 
paints of all kinds gradually grow darker with age from the oxidation 
of the oil, this oxidation progressing until either the paint cracks 
and falls off as a scale on any mechanical disturbance, or is washed 
away in the process of cleaning or by the action of storms. Chromate 
of lead, bichromate of potash, the chlorates, magnanese dioxide, 
red lead, and a number of other pigments also possess tliis oxidizing 
power to a great degree. 

Fifth. — The action of light. The action of light as a bleaching 
element is well known in almost all fields of human industry; but 
the chemical changes which occur between the pigment and the 
liquid are not well understood, this action being furthermore compli- 
cated by the different temperatures to which the coated surface 
may be exposed, and aided by the effects of sea air or fumes from 
various manufactories. We know that certain pigments fade upon 
exposure, whether applied to metallic or other structures. The 
pigments which contain organic coloring- matter from coal-tars, 
dyewoods, etc., fade more rapidly than those which have a metallic 
base; but it has never been established that the bleaching of the paint 
in all cases detracts from its durability. 



PEELING OF PAINTS. 11 

Siaih. — Peeling. Paints vary greatly in their power to adhere to 
either metallic, wooden, or other surfaces; notably zinc white, which 
peels under almost any condition or from any surface to which it 
may be applied. There is no other pigment which possesses this 
property in so marked a degree, and it is dijKcult to assign any reason 
why it should peel so badly. A possible cause is that the zinc white 
combines with the oil used in the paint and forms one of the com- 
pounds known as metallic soap, this particular one being zinc soap, 
a hard, brittle, non-adhesive substance, easily removed by mechanical 
injury, water, and in the process of cleaning, etc. Galvanized iron 
possesses the property of causing almost any paint applied to its 
surface to peel; in fact, it is one of the worst substances to cover 
with a pigment in a satisfactory manner. Experiments made by a 
leading railway company in the United States, in which a number 
of the best pigments in use by that company for all descriptions of 
railway work were tried upon galvanized-iron car-roofs and other 
galvanized work, cornices, etc., showed at the end of three years 
that but one of the list was in any manner satisfactory, and this one 
was a patented compound with bisulphide of carbon as the vehicle. 
Ordinary trade colors are of the most unreliable natiu^ when applied 
to galvanized iron exposed to the trying conditions of railway service. 
Various reasons have been given for this peculiar action of paint 
upon galvanized iron. One of the most plausible is that the use of 
sal-ammoniac in the process of galvanizing causes the formation of 
a thin film of the basic chloride of zinc on the surface of the metal 
being galvanized, which material, being of a hygroscopic nature, 
acts as a repellant to prevent the close adherence of the paint to the 
metal, and the pigment dries as a skin over it. Sheet zinc does not 
hold some kinds of paint. Sheet lead also is difficult to cover, and 
paints which take tin and lead will not always adhere to zinc. As a 
general rule, the strong oxide paints take these metals better than 
talc, ochre, and the earthy pigments. No positive general statement 
can be given, and the problem of the adaptability of paint to a metal 
to pK»vent peeling still needs study for each application. Paints ap- 
plied in cold weather, and which are exposed to a frost while drying, 
will always peel, unless the paint is warmed to about 120 degrees F. 

Another fruitful cause of the peeling of paint is when the several 
coats are successively applied before the foundation or preceding 
coat has thoroughly dried, the result being that the liquid in the 
outer or last applied coats, softens those previously applied. The 



12 PEELING OF PAINTS. 

reeulting mass, containing a notable amount of the more volatile 
el^nents of the liquid, beginning to dry from the outside surface, 
forms a thin but hard or vitreous surface which retards the further 
evaporation of the volatiles and prevents the access of oxygen from 
the air, which is necessary in the process of drying. If the surface 
thus covered has been painted while at a low temperature or during 
a damp or foggy atmospheric condition, and soon after there is a 
marked rise in the temperature or a fall in the hygroscopic conditioB 
of the atmosphere, then the paint b liable to peel at once, or soon 
after the change. This effect is hastened where the coating is a 
heavy one, or one hard to spread by reason of the earthy or mert 
substances in the pigment, or if benzine has been used as a drier. 

As a general rule, the more substances that enter into a paint, 
either as pure pigments, inert substances, or in the composition of 
the liquid, the more liable it is to peel. A small amount of fish or 
animal or non-drying vegetable oils, though oxidized by the addition 
of metallic salts and used in connection with linseed or other siccative 
oils, also hastens and provides for the certainty of the peeling. 

The peeling of paint from wooden surfaces is very common, 
particularly if applied on unseasoned lumber that contains moisture 
and air in the cellular formation of 
the wood as shown by the cut. The air 
and moisture in the cells expand upon 
a slight rise in temperature, and in thdr 
efforts to escEpe through the dried paint- 
skin, push it up in the form cf blisteiB 
that contain the condensed moisture, 
and results in the peeling of the paint 
in blisters or in strips. 

A pigment composed of a number 
of substances, the difTerent materials of 
which by themselves would form the 
basis of a good paint, when combined 
together with the liquid, necessarily 
undergo a different chemical action than 
^:^l2SS't£'^S oi ">e Beveml ™en,b.n, of the pigmert 
paiDt. would have done had they been used 

alone. This chemical action is furthermore complicated by the combina- 
tion? going on in the liquid, which, formed of a number of different 
elements that act and react upon one another, and mixed with the 



PAINTS, DESTRUCTION BY CLEANING. 13 

heterogeneous pigment, develop a series of chemical actions in the mass, 
the weaker element of which, either the mineral or the organic, is the 
first to break down or change, the decay of which hastens the decom- 
position of the others and releases the bond between the paint and the 
surface over which it is spread, and the peeling process is effected. 

That these chemical changes exist in the above stated case cannot 
be denied, but have not been well accounted for. The fact remains, 
however, that certain paints peel, and though analysis of the peeled 
portion may reveal nothing to indicate the reason for the peeling, 
it is seldom possible to get a sample of the original paint appUed, to 
compare its constituents with the peeled sample, and the cause is 
relegated to the hidden drawer of the paint-ehop, near which some 
scapegoat can be found to bear the burden of failure. (For other 
notes on the peeling of paint, see Index.) 

Seventh. — Destruction by cleaning. This cause of the deteriora- 
tion and destruction of paint applies more particularly to wooden 
structures, railway cars, and kindred objects, than to those of a 
metallic character. It may be sufficient to say we do not wash 
down an iron bridge, roof-truss, or steamship, with a view to its pre- 
senting a clean face for inspection and painting. Almost all the 
binding materials of dried paints and varnishes are more or less acted 
upon by caustic and carbonated alkalies, and but little of the soap 
in the market is free from these substances. The detergents sold 
for cleaning are all mixtures of sal-soda and caustic substances with 
lime, pumice, and other inert materials, and the more effective they 
are for removing dirt, the better they are for the destruction of the 
paint. If, in the economy of domestic household matters, two re- 
movals are equal to one fire, then it may be cited with equal force 
that two good scrubbings with any washing compound, and most of 
the soaps of commerce, applied with a stiff brush, will be equal to the 
next painter's bill to restore matters to their pristine state. Aside 
from the element of cost, it is no doubt the better practice, so far 
as the ultimate preservation of any metallic structure is concerned, that 
it should be washed clean with some of the detergent compounds of 
the day, in a very weak solution to remove the dirt, then sponged 
with a liberal amount of clean water, then be allowed to dry thor- 
oughly before the new paint is applied; but I must confess as an 
engineer, that the above method of painting is rare, and that the rule is 
for the paint to be put on regardless of cleaning the old coat, and, 
like Charity, trust it to cover the sins beneath. 



14 



WATER-TESTS FOR PAINTS, 



Eighth. — Water, The destructive action of water upon paint 
applied to structures of any material, either upon their internal or 
external surfaces, is very strong, and will rank next in destructive 
qualities to the detergent soap and scrubbing-brush. Inside painting 
lasts longer than outside, principally because it is less exposed to 
the action of water. Direct experiments show that dried linseed 
and other siccative oils, without pigment, are not resistant or 
water-repellent. When the oil is well dried, the appUcation of water 
always causes the oil to assiune a shrivelled appearance, showing 
that it has absorbed moisture and expanded, and disintegration has 
commenced. If the exposure be long continued, the whole coating 
of dried oil will slump away from the surface over which it is spread. 
Rain-water, from the sensible amount of ammonia that it carries, 
increases this destructive action on the dried oil, and the slow wasting 
away of good paints containing pigments best known to resist aging 
influences, and which have been hardened by time, can be attributed 
to this action. 

The ordinary test by master painters, of the ability of an oil or 
paint to resist moisture is to coat a surface, usually of glass, and 
when well dried, to immerse it in water for a few hours and note the 
changes in color and integrity of the paint. 

Dr. Dudley's experiments for the Pennsylvania Railroad, on the 
action of water upon paints, are interesting from the care which was 
exercised in making them and recording the results. Several sam- 
ples of a paint designed for use upon cars and wooden structures 
were made with raw linseed-oil and a very small amount of japan ; the 
same liquid being used for all the samples with varying amounts of 
pigment, all the proportions being by weight. Two coats of these 
paints were spread upon glass, and allowed to harden for two to 
three weeks. These samples were then placed side by side, and a 
small portion of the surface of each covered with a globule of water. 
This globule was covered to prevent evaporation, and then allowed 
to stand for twelve to fourteen hours. 

No. 1 was the linseed -oil and japan alone. 



2 
3 
4 
5 
6 
7 



same liquid 90 parts, pigment 10 parts. 



it 



n 



li 



n 



tt 



80 




20 


70 


. 


30 


60 




40 


50 




50 


40 




60 






bjO 

5) 






WATER-TESTS FOR PAINTS. 15 

When the proportions are higher than liquid 40 parts and 60 of 
pigment, the paint will not spread well with a brush if the liquid is 
linseed-oil and the pigment has the specific gravity of ordinary oxide- 
of-iron paints. 

At the end of the period named, the behavior of the samples 
was as follows: No. 1 coating was found to have cleaved off the 
glass and had become shrivelled wherever the water had touched 
it. Apparently the dried Imseed-oil had soaked up water, much as 
a sponge acts as an absorbent. On allowing the water to evaporate, 
the coating dried down again, but not uniformly, and was apparently 
weakened in texture. 

No. 2 showed the same characte istics. 

No. 3 showed the same, but in a less degree. 

No. 4 did not cleave off the glass, but showed where the water 
had stood. 

No. 5 showed a spot in the same way, but in a less degree than 
No. 4. 

Noe. 6 and 7 showed but very little action. 

It can be noted that here linseed-oil dried for some two months 
absorbed less water than freshly dried oil, while very old dried oil 
lost this absorbent quality and became almost water-repellent. 

To successfully design a paint which will resist all of the previously 
named destructive agencies, is a difficult matter. The field is an enor- 
mous one to cover and but little positive knowledge has yet been 
obtained, though the investigators and experiments have been legion, 
and the literature on the subject embraces volumes. Time is an essen- 
tial factor in the test of the qualities of a paint, and if the experimenter 
is required to wait five or ten years to determine the merits of any 
paint, or what effect a slight modification of the proportions has upon 
any one or more of the eight destructive agencies heretofore stated, 
a life could be spent and possibly no conclusion reached. 

Experiments are numerous in the field of designing a water-proof 
coating to be applied over the pigment which has been found to pos- 
sess the most preservative qualities, independent of the water-repellent 
features, but the goal is not yet reached. How effectually a thin 
coating of the proper material can protect the surface of a paint 
which it covers, can be seen in the lettering of old sign-boards, which 
is perhaps an example of the most durable paint of which we have any 
record. 

This protective effect is explained by the well-known fact that 



16 PAINTS, WHAT IS REQUIRED, 

lampblack is one of the best wcUer-repeUents known, that it is practi- 
cally indestructible by oxidation or acids, and being per se of an oily 
or greasy nature, when mixed with a pure oil (linseed in these cases), 
and being in a measure elastic, it has eflfectually preserved the surfaces 
and not allowed the water to reach the underlying coats of white lead. 
Having set forth the general character of what a paint should be 
for the purpose of protecting structures from decay or corrosion, 
and having indicated the most effective causes which provoke or 
promote the destruction of the object and its protector, it may not 
be amiss to speak more definitely upon those materials which enter 
into paint compounds which yield the best results in general practice. 
These results are based upon the experience thus far at hand as 
recorded or accepted data, and not the h3rpothesis of some person or 
persons whose single or joint lives may be too short a period, as com- 
pared with the life of the structure they are striving to protect from 
decay, to realize the meritorious features of their experiment. 



CHAPTER II. 
paints: statistics and general character. 

There are in the United States at the present date (1903) about 
420 firms engaged as manufacturers ancl compounders of pigments, 
pastes, and paints of all grades, representing a yearly output equiva- 
lent to about 90,000,000 gallons of mixed paints, that cost not far 
from $65,000,000. 

This represents about 570,000 short tons, and would cover with 
one coat 900,000 acres or 1400 square miles of surface, requiring 
50,000 painters to spread it. 

The following details are the average amounts of the principal 
pigments used in the United States for the years 1898 to 1902: 

Iron oxide, 23,500 short tons. Value 110.75 to $11.00 per ton. 

" " 7,000 " " " 9.00 for mortar colors. 

White lead ground in oil, 85,100 short tons. Value 5.25 to 5.50 cents per pound. 

« " dry 25,100 " " " 4.70 to 4.90 " " " 

" " iniported 300 to 700 tons. 

Red lead 11,100 short tons. " 6.30 to 5.50 " " " 

" " imported 400 to 800 tons. 

Litharge 11,000 short tons. " 6.30 to 5.80 " " " 

" imported 40 to 350 tons. 

Orange mineral 10,200 tons. " 7.25 to 7.50 " " " 

" " imported 500 to 700 tons. 

Zinc oxide 40,200 tons. " 4.00 to 4.25 " " " 

" " imported in oil 16,000 " Dry, 260 tons. 

Flake graphite 1450 tons J 

Amorphous graphite 2500 " > Value 6.50 to 6.25 " " " 

Acheson's " 30 " ) 

Imported graphites of all grades average from 10 to 12 times the amounts 
produced in America. 

Ochres of all grades were produced in 13 different States, Pennsylvania fur- 
nishing over one-half of the entire output of 14,200 to 14,500 short tons. Value 
16.50 to S7.00 per ton. 

Imported ochres, 7,700 to 8,000 tons. Value $7.70 to 17.90 per ton. 

Spanish brown, principally from Maryland, 600 to 650 tons. Value $17.70 
to $18.00 per ton. 

Approximately, the United States produced 248,600 short tons of the above 

pigments, paints, and pastes, against 53,300 '' " imported. 

17 



18 PAINTS, GENERAL CHARACTER OF. 

What proportion of these amounts were really applied for the 
preservation of metallic structures on shore or afloat, it is diflScult to 
determine; but one-fourth part may be taken as the yearly allowance 
to cover the efiFects of corrosion in progress in some degree in about 
every metallic structure that meets the eye, and may be considered 
as the annual contribution to the coffers of corrosion. 

The general tenor of paint-trade literature would lead the layman 
to infer that each one of the above noted 420 firms was the right 
and only one that could or did furnish the special and imperishable 
paint that he was in search of. The customer who is in search of 
facts as well as paint will find some of the former in Chapters V, 
XXXII, XXXIII, XXXIV, XXXV that may guide him in selecting 
the latter. 

The greater part of the mixed pastes and paints of the day are 
adulterations, and are presented to the public in these forms the 
better to conceal the actual composition of the pigments and to save 
oil; also to disguise the quality of the vehicle, as in the form of a 
paste or paint it requires chemical skill and time to analyze a sample 
of either. This, while applying in general to the mixed-paint house 
colors, does not exempt large quantities of mixed paints sold exclu- 
sively as ferric protective coatings. 

There are at the present day as pure brands of linseed-oil, red 
and white lead, lampblack, and other pigments manufactured as 
any ever made. Possibly they are better on the average than 
those made one hundred years ago; but there are more that are a 
great deal poorer, and rendered more so by adulterations of the 
most barefaced character. There is a great advantage in the use of 
prepared paste, as the quality of the vehicle required to bring 
it to paint can be positively known, also the driers used, and what 
amount of these is necessary to meet any condition present at the 
time and place of applying the coating, — details that in most cases 
cannot be known in advance or by the paint-compounder, unless, 
as is too often the case, he makes but one kind, and fits it for the con- 
templated duty by the difference in price and the gullibility of the 
customer. 

There are many reputable and responsible manufacturers of, 
and dealers in mixed paints, who will and do give a statement of 
the materials they use in a brand of mixed paint and the reason there- 
for. But they do not sell pure linseed-oil, under a fancy trade-mark, 
for 19 cents a gallon, nor a paint for 40 cents a gallon, that if the 



PAINTS, GENERAL CHARACTER AND COST, 19 

given materials were only approximately pure, would cost nearer a 
dollar. Neither do they expect it to be spread with a stiff, hard 
brush to enable it to cover a large area as a recommendation for its 
cheapness and superiority. 

It costs as much for labor, brushes, scaffolds, and other items to 
spread a poor paint as a good one. On railway bridges, viaducts, 
and structural ironwork painted in situ, it costs for the painters' 
labor about twice the cost of the paint and in many cases four times 
as much, — depending upon the character and amount of scaffolds 
or ladder-work. 

This assumes that a reliable paint is used that costs about a dollar 
a gallon, that will cover from 300 to 400 square feet of surface for 
the first coat, and from 500 to 600 square feet for the second or a 
repainting coat. In the latter case the surface covered may be 
less, or the same as for the first coat, all depending upon the labor 
of scraping or the condition of the surface of the old coating, whether 
scraped or not. Obviously, the claim that a paint can coat 1000 
square feet of surface or more, and prove as durable as those covering 
less surface, as above, is not sustained in practice, though it is always 
possible to get a doctored result with any paint, good or bad. Paint- 
films — that is, the oil covering the atoms of the pigment — are only 
from y^iy to juVt ^^^^ ^^ thickness, whatever the size of the pigment- 
atoms. It stands to reason that a thick coating of the vehicle will 
better protect the pigment-atom than a thin one. If the pigment- 
atom is susceptible in any degree to atmospheric influences, it will 
be less affected with a heavy coating of the vehicle than with a thin 
one. A thin coating usually implies that the oil has been reduced 
in density to render it easier to spread, and to be spread over a larger 
area, by the use of a larger quantity of solvents, either turpentine or 
benzine, than is necessary with any good quality of either raw or 
boiled linseed-oil. 

Red-lead paint, from the large amount of oil in it and its great 
specific gravity, spreads over a large area, and it is these features 
that cause it to run or crawl on vertical or slightly inclined surfaces, 
particularly in the first coat. 

A like result follows the use of flake-graphite pigments. The 
atoms of this variety of graphite, on account of their smooth surface 
and low coeflficient of friction, appear to slide around in the vehicle 
before it dries enough to retain them in position when spread. The 
silica and barytes frequently mixed with such pigments to give a 



20 ESSENTIAL ELEMENTS FOR SECURING GOOD RESULTS. 

frictional resistance to overcome this gliding are almost as non- 
absorbent or repellent of the oil as the flake-graphite atom, and 
have a greater specific gravity to crowd them downward. 

While the following excerpts * and the author's views are in the 
main more applicable to railway bridges and structural ironwork 
than to house-painting or the many minor ferric or composite surfaces 
that require painting, possibly, more for appearance than protection 
from corrosion, yet it is quite apparent that there is too much poor 
paint spread in both cases. The frequent failures or inferior results 
of all kinds of paint applied to all classes of structures can be attributed 
to, not only the paint and the manner and time of appljdng it, but to 
the improper preparation of the surface to be covered. 

This is particularly the case in regard to bridge and structural 
painting. Particular emphasis is laid upon this point by every author- 
ity and writer on the subject in the technical literature of the master 
painters and engineering associations' debates and reports. Every 
specification for painting, bristles with clauses prescribing what shall 
or shall not be done, and still the fact remains that there are more 
failures than even indifferent successes, especially on work painted 
at the shops before shipment. The causes for the irregular and indif- 
ferent results are not difficult to ascertain. They are the improper 
application of the paint to dirty, greasy, moist or chilled, rusty or 
mill-scaled surfaces. No marked improvement in these imcertain 
results can be had until the same importance is attached to the "paint 
question," not only on paper, but in the actual supervision of the 
painting in all of its stages, as is given to the minutest construction 
details. 

How carelessly this* essential is generally performed even by 
engineers in charge of large bridges and other structures is seen in 
the instance of an inspecting engineer who reports visiting the work- 
shops to observe the condition of the metal after being manufactured 
into structural work and during the painting process. "The metal 
had been housed, but there were many rust-spots on the web-plates, 
also on the angles, which were covered with scale. The metal was 
being cleaned by puMy knives and whisk-brooms. Steel brushes were 
sometimes used (presumably as long as the visitor was present). // 
there was anything unusual in this method of cleaning at the shops it 
was on the part of thoroughness, (The italics are the author's.) After 



* Excerpts from Engineering NewSj June 6, 1895. 



PAINTS, ENGINEERS' RESPONSIBILITY FOR A GOOD RESULT. 21 

cleaning; the plates still showed thin yellow rust-spots, that showed 
plainly, but of a darker color after coating with oil. The oil was 
scraped from some rust-spots under the oil on dry girders in the 
yard, and the yellow color of rust, so often found, was developed." 

It is to be regretted that this engineer's views of what constitutes 
a thorough preparation of the ferric surface for its coat of paint is 
not an exception, but the rule in more than nine-tenths of the struc- 
twrsA manufacturing establishments. Notwithstanding their claims 
to pre-eminence in their profession, they have yet to learn how to 
protect what they create; and that they are either incapable of this, 
or indifferent to it, the present condition of the ferric structures of the 
day is an unanswerable evidence. 

If the superiors do not understand the importance of the proper 
preparation of the surface to be covered, or the character of the paint 
and manner of applying it, or give them the same or more con- 
sideration than they attach to other matters of construction, it will 
be next to impossible for the inspector or master painter to enforce 
good work. It requires a more determined stand on the part of those 
in charge of this branch to ensure good work, than in any other part 
of the construction details. Until the head oflficers are zealous enough 
to care something about the condition of the work after it has left 
the shop, and the men actually in charge of the painting are given to 
understand that they will have the unquestionable backing and sup- 
port of their superiors in any stand they take against the present so- 
called practical methods of structural painting by unscrupulous 
contractors, just so long will their work show their neglect in the 
rapid progress of corrosion, that will not need scraping the surface of 
the coating to find. 

The low grade of labor available for the painters' gang has much 
to do with the generally unsatisfactory results obtained. Painting 
can be slighted and still present a creditable surface that will pass 
inspection more easily than any other branch of hand labor connected 
with bridge or structural ironwork. Painting is as hard in muscular 
requirements as light blacksmithing or the vise-work of a machinist, 
and the painter is not addicted to wasting his elbow-grease to work 
out his paint over any larger area than he can well avoid. In con- 
tract painting this element is noticeable, as there will often be 25 
to 30 per cent of "difference in the areas coated by different workmen 
upon the same job, and the eye can hardly detect the difference. 
The regular bridge inspector in charge of the work at the shop is so 



22 CLEANWO SURFACES PREPARATORY TO PAINTING. 

crowded with miscellaneous duties, that the inspection of the painting 
is usually a farce, even if the quality of the paint, the weather, and 
other conditions are favorable to secure a first-class result. 

These features are particularly apparent if red lead is the paint 
used for the shop coat, as way wwat of care in keeping it continually 
well stirred up in the paint-pot by the paddle-stick (not by the brush) 
to prevent its "setting " is almost undetectable, and the want of care 
here governs the durability of all of the subsequent coatings. The 
use of lampblack with red lead in a paint coating, while it delays 
the quick "setting" of the coating, does not prevent the rapid settling 

of the pigment. 

Probably the best results could be obtained if the man or firm 
who pays for the completed structural work appointed his or their o^\^l 
inspector to attend to this branch of the work with the distinct imder- 
standing that his orders were to be strictly enforced, and that his 
endorsement on the bill rendered was necessary before pa3mient of 
the same. This would ensure the proper preparation of the surface, 
and secure careful attention to the before-mentioned necessaries; and 
he alone could be held responsible for the final results. In general, 
railway bridges that have the several coatings of paint applied under 
the direct supervision of one of the railway company's own corps of 
engineers have proven to be better protected against corrosion than 
the structures painted either by contract or by the most prominent of 
the construction firms, who, as a rule, are more anxious to get the work 
out of the shop, than for its future fate. 

The pickling of structural iron with dilute acids to remove the 
mill-scale, as done in some classes of ship and boiler work, has met 
with many objections. These objections are primarily the cost of 
the process compared with a rush coat of something denominated 
paint. 

When pickled and brushed clean of scale, the metal mxist be copi- 
ously washed in water and then dried if possible in the sun, or artifi- 
cially in a warm room or oven, and then, whether machined or not, 
be coated with the first coat of paint. Tf a few hours elapse before 
applying the coating, the surfaces will begin to acquire the thin 
blush coating of red nist, as described in Chapter III. 

The use of the sand-blast at the final stage of the machining proc- 
esses will effectually remove the dirt and scale, but the machine-grease 
must be soaked free with turpentine and thoroughly wiped off, and 
not allowed to dry down again. 



PAINTING AT THE MILL. 23 

Both the pickling and sand-blast processes cost money, patience,- 
and grim determination to apply, but the result in having a properly 
cleaned surface for the foundation of the protective coatings has 
been proven in hundreds of cases as the only sure method to reduce 
the maintenance expense of the structure. (See Chapter XXVIII, 
Sand-blast and Pickling.) 

Many engineers are advocating the plan of having a coating of 
either boiled oil or paint applied to the iron or steel at the mill as 
soon as possible after it has left the rolls or hammer, and while the 
metal is hot. The hot part is the only part to commend. All metal as 
it leaves the rolls or hammer has a tough, thick or thin (as the case 
may be) coat of loose or partly loose scale that adheres for the time 
being, but on a short exposure to the air with a few changes in tem- 
perature, due to mill or storehouse conditions, releases its tension 
and is ready to fall off whenever handled, as in the course of loading 
and transportation. No amount of brushing that any mill employ^ 
would or could give to the metal in its hot or half-cold condition would 
remove this scale, and if the painter was present with his pot of oil or 
paint, it would get on over scales and aU, and no ordinary inspector 
could prevent it, or be in any way sure that the contract requirements 
had been complied with in regard to the removal of the scale or the 
composition of the coating. 

The mill coating is exposed during its application and drying 
to all the dirt, cinders, and sulphurous gases of the mill, which are a 
fmitful cause of decay in a dried coating of paint, and find an easier 
field in the green one. The mill-coated work is not allowed time to 
dry before being loaded for transportation, which adds its quota of 
dirt and cinders to the sticky paint. 

All the subsequent machine operations are accompanied by more 
or less lubrication of the tool, and the oil used for this purpose is the 
cheapest to be had, and in general has been used over and over again ; 
is dirty, sour, and more or less decomposed, and carries enough hydro- 
carbon to evaporate and dry down as a dirtj^ surface skin, hard to 
distinguish from the coating applied at the mill. The sequence is 
that the inspector crowded to get the work out of the shop, and if at 
all careless in the discharge of his duty, does not personally see that 
the scales, dirt, and machine-grease are properly removed. The 
painter, anxious to show a great day's labor, and as a class prone to 
scrimping everything that calls for any manual effort other than 
with his brush, and jealous of any attempt to confine him to a pro- 



24 PAINTING AT THE MILL. 

cednre at variance with what he thinks is a special function of his 
craft, hastens to get on the paint, and takes more credit to himself 
in being able to beat the inspector than to do a meritorious piece of 
work. 

Rather let the material go from the mUl or forge to the storeroom or 
construction shop, protected as far as possible from any unnecessary 
exposure to the elements. When machined, during which process 
the greater part of the mill-scale will be loosened up so as to be readily 
removed, and when the several parts are assembled in their relative 
positions ready to be riveted up for their permanent places in the 
structure, if it is to be done at the shop instead of in situ; then and 
there is the place for the inspector to determine if the several parts 
are not only properly machined, but also properly cleaned from the 
scale that has not been removed by the machining and handling. 
He should see that grease, dirt, and any remaining scale, tight or loose, 
is removed in his presence, and the first coat of the paint applied in a 
manner to meet the atmospheric conditions at the time, and use a 
quality of paint that will ensure more than a guess at the future pro- 
tective result. 

Nothing can then serve as a cloak to hide the inspector's responsi- 
bility for the result. One inspector, and one inspection at the final 
stage, is better than a number of inspectors and inspections strung 
over a chain of operations comprising months of time and hundreds of 
miles between the links. 

Many engineers advocate the use of boiled oil alone for the first or 
priming coat, applied either at the rolling-mill to protect the metal 
during its transit from the mill to the construction shop, or at the shop 
when ready to ship for erection. The general reason asi^igned for this 
practice is, that the boiled oil ''soaks into the scale and dries and pre- 
vents further tendency towards corrosion." 

This theory is absolutely without proof, from any standpoint. 
How far any oil or liquid can soak into iron or steel or the still 
harder mill-scale that forms on these metals, these Solons do not 
state. The use of such oil coatings is, in general, to conceal some 
slop-work on the part of the inspector, or constructor, at an earlier 
stage of the work than would be possible later on. However consist- 
ent and beneficial the first coating of oil may be for a wood or masonry 
surface, it has no part or parcel on a metallic one, when applied for 
the correction of the mill-scale evil. No number of these oil or even 
paint coatings will soak into and bond these scales together, or to 



BOILED L1NSEEIU)IL COATINGS. 25 

the metal surface. There are hundreds of records of the painting 
of important railway structures, where the first coat of boiled-oil 
method was used, and, in the great majority of instances, the utter 
and rapid failure of the coating, wad the extra corrosion of the struc- 
ture, could be directly assigned to this so-called method of protection. 

The weather-resisting power of an oil coating is almost nil com- 
pared with a paint, as before referred to in Dr. Dudley's experiments 
(Chapter I). If the advocates of oil coatings are so sure of its bene- 
fits as against a paint, why not make all the coatings of oil alone, no 
matter what it covers, a wire or an anchor? It will soak as far into 
one as the other. A paint coating can be applied as quickly and easily 
to any surface as an oil coat; wiU dry as quickly and as hard, and is 
in every way a better resistant to atmospheric or mechanical injuries. 

A foundation coat of oil is a direct cause of the blistering and peeling 
of the coatings spread over it. It is seldom dried enough before the 
other paints are spread over it, to ensure a close adherence to the metal 
it covers. When the subsequent coats of paint are spread, the solvents 
and oils in them sbften to some extent the underlying coat of oil, and 
a moderate heat from the sun causes the whole coating to blister or 
peel. Too much oil in a paint coating, particularly if the surplus oil 
is in or near the foundation coat, whether on a wooden or metallic 
surface, will generally cause peeling regardless of the pigment used in 
the coatings. 



CHAPTER III. 

IRON. 

Symbol, FE. Atomic weight, 66. Specific gravity, 7.77. 

Iron is never found pure in nature. Its avidity for oxygen is so 
great that it quickly forms ierrous oxide, FeO, or the protoxide of 
iron. This also is never found free, and is diflficult to obtain chemi- 
cally pure, its affinity for oxygen forming the sesquioxide — Fe^O, — 
called also the peroxide (the highest form of oxide for any metal) 
in which two atoms of iron and three of oxygen are united, or 70 
per cent of iron and 30 per cent of oxygen. 

In the latter form it is commercially knowm as iron oxide or iron 
ore, and is found in all parts of the world in all stages of purity, and 
in combination with the oxide of all the other metab in all propoi- 
tions. The color of the protoxide is a green hue changing to a red- 
brown — that of the peroxide is a blood-red. 

An intermediate oxide — the black magnetic, Fe304, three atoms 
of iron and four of oxygen = 72.4137 percent of iron and 27.5863 per 
cent of oxygen — ^is the purest oxide of iron. 

The ferric anhydride, FeO,, is not known in nature, but is sup- 
posed to be formed by fusing iron or its oxide with nitre. Its color 
is a deep crimson. 

Iron at a temperature of 230° C. (446° F.) combines freely with the 
atmospheric oxygen, becoming first covered with an extremely thin 
film of magnetic oxide, F^Oi, of a light yellow color, which gradually 
passes into red. blue, and gray color. At a white heat, iron bums in 
the air with a production of magnetic oxide, the combustion being 
sustained for some time by directing a blast of air upon the heated 
metal. At a temperature of 360° C. (680° F.), iron decomposes steam, 
fonning the black magnetic oxide of iron, Feg04 (the Bower-Barflf 
coating), and liberating hydrogen. 

CJrocus, a fine powder formed when iron-scrap, borings, or ore are 
placed in contact with malleable- or cast-iron articles in a closed recep- 
tacle, and all brought to a red heat for the purpose of annealing 

them, is anhydrous iron oxide, FcjOj, of a dull red-brown color. It 

26 



OXIDES OF IRON. 27 

is no better when used for a pigment than any natural iron ore, other 
than in its freedom from sulphur. It hydrates on exposure to the air 
or moisture to FejOj+HjO, and can be reduced to metallic iron the 
same as any iron ore. 

The precipitate formed from metallic iron when corroded under 
water is the sesquioxide or peroxide of iron, FejO,, plus three parts 
or 24 per cent of water, and is red rust, FcjOj+SHjO. It is a 
dull reddish-brown color, nearly a pure oxide, containing only such 
other metalhc oxides as the iron contained from which it was cor- 
roded. It is comparatively free from sulphur, more so than the best 
hematite ore. 

Oxides of Iran. 

If purity of an iron-oxide pigment is any factor to prevent corrosion, 
these pure oxides ought to be better than any ircn-oxide ores; but 
they are not, and plainly show that the failure of all iron oxide- 
pigments to prevent corrosion on a ferric body, or to add any resist- 
ance to the decay of the paint coating, lies in the natural inadequacy 
of a ferric pigment to resist its own inherent weakness, namely, con- 
veying excessive amounts of oxygen with a tendency to excite elec- 
troljrtic action. 

Experiments determine that bright iron placed in an atmosphere 
of dry oxygen, or of dry carbonic acid, will not rust; when put in a 
damp atmosphere of oxygen, it rusted slightly; in a damp atmos- 
phere of carbonic acid, a small quantity of white carbonate of iron 
is formed on the surface of the bright metal, but no rusting takes 
place. When, however, bright iron is placed in a damp mixture of 
the two gases — oxygen and carbonic acid — it is rapidly oxidized 
into copious excrescences of red rust. 

In the opposite direction, to prevent rusting, a strong solution of 
carbonate of soda preserved needles and steel instruments, bright 
and untarnished, after thirty years of exposure, and would probably 
do so forever 

Bright steel or iron objects remain untarnished in un atmosphere 
of diT muriate of lime, also in the dry carbonate of lime. Iron im- 
mersed in lime-water, caustic potash, and caustic soda does not rust; 
though the lyes absorb carbonic acid, they do not absorb oxygen. 

The solutions of chloride of soda, kalium, magnesium, and ammo- 
nium quicken the formation of rust the same as dilute solutions of 
acids, if free oxj'gen has access. Under atmospheric influences 



28 OXIDES OF IRON. 

that oxidize rino, lead, and copper, the layer of the oxide formed is 
measurably thick and pre^'ents any further oxidation. On the con- 
trary, iron-rust once formed on a ferric surface never ceases iie action 
so long as it is in contact with it. Rust produces rust. 

The blush of oxide that appears upon the surface of a piece of 
bright, rlean iron, such as is Mt from the action of the sand-blast or 
of a grindstone, forming after a few hours of atmospheric exposure 
and can be wiped off by the hand; or the piece of red rust fiom the 
iron-ecrap heap; or the scales from the bottom or frame of an iron 
ship; one and all are the peroxide of iron, FejO,, plus three parts or 
24 per cent of water. When calcined to drive off the water, they 



become precisely the same ferric oxide contained in any iron ore, and 
wiD reduce to metallic iron or grind to a pigment the same as any 
iron ore, however they may be designated or juggled with trade- 
names. 

In liie corrosion of iron from any cause, for every 8 grains 
in weight gained by the iron, 46 cubic inches of hydrogen weighing 
1 gnun are set free. With each ounce of gain in the weight of the 
iron, 251S.625 cubic inches of hydrogen (=1.4558 cubic feet) are 
evolved, weighing i ounce. Every pound of the oxide of iron re- 
quires the evaporation and dissociation of .2962S of a pound of water, 
representing the energy of about 1 pound of coal. 



OXIDES OP laON. 29 

The magnetic oxide of iron is the richest of the iron ores in metallic 
iron— 72,414 per cent. It is non-corrosive, and, in the form of black 
titaniferous sand, found on the seacoaat in many parts of the world, 
exposed to sea-water and other sources of oxidation and friction, has 
remained unchanged for thousands of years, Its use for a pigment 
is not satisfactory, on account of its black color and the difficulty 
of grinding it. 

Specular iron ore is also but little affected by oxidation, and is a 
nearer approach to a definite compound of iron and carbon than any 
other known ferric subetance: iron, 94.85 per cent; carbon, combined 



and graphitic, 3.S0 per cent; silica, manganese, sulphur, and phoe- 
phorus, 1.65 per cent. It is an anhydrous ferric oxide, found in 
Nova Scotia, in the Isle of Sicily, and other parte of Europe, 
where mines of it have been worked for 3000 years. Its black color 
and hardness prevent its use as a pigment, though its resistance to 
corroeioQ is almost equal to that of the Bower-Barff surface. 

He clay-iron ores from the coal measures, the spathic, bog, and 
many other iron ores, contain a very small amount of iron, and so 
laige an amount of silica and other mineral substances that they are 
too refractory for smelting. They are not used for pigments for the 
same reasons that attend the magnetic and specular ores. 



^ HEMATITE ORES. 



Hematite Ores. 



The red and brown hematite-iron ores, composed of the sesqui- 
oxide of iron— 70 per cent metallic iron and 30 per cent oxygen, plus 
water, plus variable percentages of mineral substances, plus carbonic, 
sulphuric, and phosphoric acids (see following analyses)— are the 
principal metaUurgical iron ores, also those used for the production 
of pigments under the name of iron oxides, Fe^Os. This chemical 
symbol, name, and product is subject to more commercial jugglery 
to meet trade requirements than any other pigment m use. It is 
made to cover all sorts of combinations diveree in composition and 
character, supplemented still further by quantities of so-called inert 
bodies, more unstable than those in the ore with which they have 
been brought into forced relation. 

These hydrated ores when calcined to expel the moisture-sul- 
phuric, phosphoric, carbonic, and other erroneously supposed easily 
evaporated acids— become the anhydrous or supposed neutral, FcjOg, 
plus about 2 per cent of water in a combined form, plus the acid 
elements that frequently amount to 2 per cent. Only 30 to 50 per 
cent of the sulphuric and phosphoric aeids are dispersed in the com- 
paratively low heat of the roasting process, and are not wholly con- 
sumed in the high heat of the blast-furnace, as manufacturers of 
metallic iron find to their annoyance. 

The lime, magnesia, alumina, silica, manganese, and other oxides 
whether in a combined state in the ore, or as added free substances at 
the time of roasting by calcination, become caustic and hygroscopic 
and when ground to a pigment form, absorb moisture from the atmos- 
phere, slack, changing their character again more or less to a floccu- 
lent or a powdered state. They do not bond in the slightest degree 
to the oxide of iron or base, are no more connected with it or to each 
other — except in a haphazard arrangement of their disrupted, sepa- 
rate natures — than the same substances would be if collected from a 
sand-bank. 

No mechanical process connected with their incorporation into 
a pigment or paint can arrange them in sequence, or in any order 
where either substance can be supposed to protect its neighbor or 
even itself from any disturbing cause. Then- excited, unstable 
condition and close association in a finely powdered form in the 
paint, render them only the more susceptible to catalytic action 
among the several substances of the pigment. This action soon de- 



IRON ORES, QUALITIES OF. 31 

stroys the weakest of them by electrolysis, set in action by their 
association as positive and negative electrical elements, or by the 
catalytic power of nearly all finely powdered substances to condense 
moisture and gases from the atmosphere, which the porous nature of 
the paint coating readily absorbs. If sulphur is present in either 
the pigment or vehicle in any recognizable quantity (as it nearly 
always is), it furnishes an additional excitant for the electrolytic 
action. This electrol3rtic action is further intensified by the unequal 
composition of all iron ores, whether roasted or not. The process of 
roasting — always an uncertain one — does not affect the ore equally. 
Lumps improperly roasted, or from their composition affected dif- 
ferently by the process, are difficult to detect in the hasty and gener- 
ally poor assorting or picking-over the ore receives before pulveriz- 
ing. The same uncertainty in the composition and assorting attends 
the unroasted ores. 

In the pulverizing process there are many larger and harder 
particles of the ore that would not pass a No. 50 mesh sieve, if 
the pigment were bolted (as it seldom is), and would much less pass 
a 100 mesh, to which size all pigments should be reduced. 

The finer the pigment the more thoroughly will it incorporate 
with the vehicle and protect itself and the surface covered. The 
destruction of any particle of the pigment will not render the coating 
so porous as when a larger atom is removed to permit access for the 
atmospheric moisture and gases. These lumps act as centres to 
determine the corrosive action, and in a measure explain the erratic 
action of all iron-oxide coatings. In nearly all rust-spots, one or more 
of these hard particles will be found, and particularly so wherever 
pitting has commenced. 

The brown hematite ores are claimed to be practically free from 
sulphur, therefore the best for a pigment; but the best brands of 
this variety of ore prepared by any one of the many manufacturers of 
unroasted iron-oxide pigments have not proved to be in the slightest 
degree any more reliable in composition, or any better protection 
against corrosion — whether used as a straight paint or mixed with 
adulterants — than those prepared from the red hematites. (See 
following analyses of both pigments as used in commercial paints.) 

The dirty purplish-brown or lifeless color of the brown hematites, 
even when freshly applied and aided by the gloss of oil, is not agreea- 
ble to the eye. Their ability to carry a large amount of uncombined 
substances, as inert pigments which add no quality to the paint as a 



32 



IRON ORES, ANALYSES. 



protective agent, and the low cost of the whole line of iron-oxide 
pigments; are the great inducements for their production and use 
for ferric coatings. 

ANALYSES OF IRON ORES. 

By Vabious Analysts. 
(Specific gravity from 5.33 to 4.85.) 





Brown Hematite (Limonite), 
24 Different Ores (Hydrated). 


Red Hematite. 8 Different 




Oree (Anhydroiin). 




PeroentagBB. 


Av'ges. 


Peroentaces. 


Av'ges. 


Ferric oxide from 


... 90.05 to 32.76 


59.54 


98.71 to 66.55 


89.13 


Ferrous " " . 


... traces " 10.64 


1.22 


traces " 1.13 


00.16 


Manganous '' " . 


... 3.06 " 0.05 


0.87 


1.13 " 0.10 


0.31 


Alumina " " , 


... 27.95 " 0.05 


4.48 


2.79 " 0.06 


0.82 


Lime " " , 


... 0.06 " 27.72 


4.20 


9.40 " 0.37 


1.97 


Magnesia " " , 


... 0.17 " 10.21 


1.30 


1.39 " 0.08 


0.42 


Silica " " . 


... 63.52 " 0.79 


13.90 


8.90 " 1.00 


5.77 


Carbonic acid " . 


... 0.16 " 18.45 


3.83 


5.73 " 0.78 


1.35 


Phosphoric " " . 


... 0.06 " 3.17 


0.63 


1.02 " traces 


traces 


Sulphuric " " , 


... traces " 0.28 


0.03 


1.31 " " 


« 


Iron pyrites " , 


... " '* 0.30 


0.03 


traces 


It 


Water combined and b 


y- 








fiTOSCODlC 


.. 18.60 " 6.60 


10.25 


2.12 " " 


ti 








Percentages of metallic 


iron 63.04 to 24.09 42.45 


66.10 to 47.47 


62.37 


11 samples averaged. . . 
13 ^' " ... 


.... 52 .65 per cent of iron. 


1 


.... 33.81 " " " " 


Only one sample was 
below 66 per cent of 










ux)n. 



There is a wide difference among these comparatively few and 
better quality ores, selected from many hundreds of ore-beds, on 
accoimt of their purity and high percentage of metallic iron. 

Many other mines furnish ores that are worked for the other 
metallic and chemical substances they contain, as nearly all the other 
metals are found associated with iron. All iron mines are noted 
for the variable quality of the ore taken from the same or from the 
adjoining bed, or from different parts of the same vein in each mine. 
The hematites are not exempt from this feature, whether used metal- 
lurgically or for pigments. 

Though an analysis may show an iron ore to be good for metallur- 
gical purposes it does not follow that such an ore is suitable for a pig- 
ment, however much the hematites may be desired as such, on account 
of their supposed freedom from sulphur or ease in grinding them 
unroasted. 



IRON-OXIDE PIGMENTS. 33 

Ores containing 40 to 60 per cent of the sesquioxide of iron and 
30 to 50 per cent of silica have not proved to be any better protection 
against corrosion than those containing 80 to 95 per cent of the ses- 
quioxide. This will be apparent by reference to the composition of 
the iron-oxide pigments given in the tests of commercial paints, 
Chapter XXX. 

IrorirOoMe Piffments. 

The red hematites furnish a brighter-colored oxide than the brown 
hematites, whether roasted or not. The small amount of sulphur 
in the red ore develops in the process of roasting the dull-red color 
into a brighter red, simulating the Venetian and Indian reds so desira- 
ble to produce, but not always possible to get without doctoring the 
furnace product subsequently with substances more complex and un- 
stable than the iron oxide itself. (See inert pigments, Chapter XVIII.) 

The roasting process is a sensitive one. A few degrees of higher 
or lower temperature, or a little difference in the period of exposing 
the ore to it, or in the manner of cooling down the furnace, cause a 
great range in the color. The more sulphur in the ore the brighter 
the color. 

There are from 10 to 20 per cent of moisture and carbonic acid 
in all iron ores as they come from the mines. If these are not driven 
off by roasting, they will not be dissipated in the pulverizing, and 
will be carried by the pigment into the mixed paint to its detriment. 
The use of an uncalcined iron-ore pigment is a long step toward an 
early corrosion of the ferric body over which it is spread. 

The following analysis of an iron-oxide pigment made from a 
special red hematite roasted ore, one of the oldest and best known of 
this class of pigments, and the use of which as a special brand is 
probably greater than all of the other brands of iron-oxide pigments 
in the world, is of interest for comparison with an unroasted ore pig- 
ment: 

manufacturers' analysis. 

Peroxide of iron 52. 11 per cent. 

(Equivalent in metallic iron, 36.477 
per cent.) 

Silica combined 46.03 " " 

Lime 0.23 " " 

Moisture 1.59 " " 

Loss 0.04 " " 

100.00 " " 



34 IRON 'OX IDE PIGMENTS. 

Certainly, this is a pigment that should show a good result on 
either wood or iron surfaces, if there is any protective value in iron 
oxide. However, notwithstanding its ahnost uncontested use for 
over thirty years, on account of its low cost, agreeable color, and 
much lauded protective virtues, it proved so unsatisfactory for both 
wood and iron coatings that various railway companies — the largest 
consumers of paints — ^have reduced the 50 per cent of this peroxide 
of iron admissible in their mixed-color paints to 25 per cent. But 
this change has not resulted in any marked improvement in the 
protective qualities of the paint when applied to ferric bodies, nor 
are better results apparent upon wooden surfaces. 

The following analysis is of a brown hematite unroasied iron ore. 
With a number of other brands of similar composition, it has been 
largely used by construction and railway engineers upon hundreds 
of the most important ferric structures in the country, whose serious 
corrosion, after but a short period of exposure, led to a special exam- 
ination and report on their condition to the engineering firms responsi* 
ble for their erection and condition: 

Peroxide of iron 03 . 04 per cent. 

(Equivalent in metallic iron, 
65.128 percent.) 

Silica combined 3.28 " 

Alumina combined 2 . 385 " 

Lime and magnesia 0. 66 " 

Organic and volatile 0. 42 " 

Sulphuric acid 0.03 " 

Moisture and loss 0. 185 " 



100.00 '' 

This is a high-grade metallic iron ore comparatively free from sul* 
phur, and whose merit as an anti-corrosive pigment was greatly com- 
mended (by the manufacturers) for a straight paint free from the 
usual class of inert adulterants. But its protective results as detailed 
in the said report were not better than the other adulterated oxide 
pigments, or the other coatings of mixed or unknown composition. 

The following is an analysis of an iron-oxide pigment that is 
reported to have given very good results, being among the best of 
that class of compounded iron-oxide paints:* 

* Messrs. Hunt & Clapp, chemists and bridgo inspectors, Philadelphia and 
Pittsburg. 



IROX-OXIDE PIGMENTS. 36 

Iron oxide 26.72 per cent j =^^ ^^-JO P^.^ ^^* 

^ { of metallic iron. 

Carbonate of lime 30.19 " " 

Sulphate of lime 14.05 " " 

Clay and sUica 19.90 " " 

Alumina 8.18 " " 

Ma^esia 0.52 " " 

Water and organic matter 0.44 *' " 



100.00 



it u 



Additional examples of iron-oxide paints and their erratic action 
— both mixed and straight pigments — will be found in the article on 
paint tests, Chapter XXX. 

It is claimed that iron-oxide pigments, being the peroxide of 
iron, are incapable of further oxidation, and when ground with the 
vehicle are indestructible, and their capacity to condense atmos- 
pheric moisture and gases ceases. This is true as long as the thin 
film of the dried vehicle — only -5^^ to y}^ inch in thickness — 
remains in place on the external surface of the pigment atom, and 
no longer. The same causes that remove this film will affect the 
other part of the vehicle, in which the pigment atoms are im- 
bedded. The vehicle, passive of itself to condense atmospheric 
moisture or gases, is porous and absorbent; and passes them on to 
the point where their decomposing action can take effect, if not on 
the iron-oxide atom, then upon the less resisting mineral substances 
associated with it as a pigment. 

With the possible exception of silica and bar3rtes, all of the 
so-called inert substances, usually mixed with iron-oxide pigments, 
are porous and absorbent of the vehicle and gases that reach 
them. The protection that these inert substances receive from 
the oil is no greater than the oil affords the iron-oxide atom, if not 
less, owing to the unreliable character of their composition naturally. 
If they have been mixed with the iron ore during the process of 
roasting, they are rendered more unstable, and readily pass to a 
lower plane of resistance, as mentioned before. 

It may be questioned whether iron oxide is incapable of con- 
densing moisture or gases. It induces and promotes oxidation in 
all organic bodies with which it is brought into contact, and is used 
in the process of boiling oil to increase the catalytic power of all 
siccative oils to absorb oxygen from the air. Its use with raw linseed 



36 ironjoxide pigments. 

and other oils is to absorb the glyceride element that, unabsorbed 
or unchanged, in all fatty oils delays the drying process, condensing 
the atmospheric moisture and gases that act below the external 
film of the drying oil, thus laying the foundation for a blister with 
subsequent corrosion of the coated surface. 

The power of iron oxide to absorb the glyceride is about two- 
thirds that of red lead. If the iron-oxide atom is insensible to the 
presence of sulphur that may be presented to it in any form, the 
other associated mineral substances and vehicle are not, and a 
very small percentage of any acid will set in motion the electrolytic 
action so fatal to ferric substances. 

As a class, the inert pigments are electro-positive to the iron- 
oxide atom, and are the first to be affected by any electrolytic action 
inaugurated by their presence in the paint. 

The iron-oxide atoms are electro-negative to the ferric surface 
over which they are spread. 

In iron-oxide and zinc-oxide mixtures, the iron atom is electro- 
negative to the zinc atoms, which are quickly destroyed. If any copper 
or copper oxide is present in an iron-oxide pigment, the iron oxide 
is electro-positive to the copper and is destroyed. 

Mallet's experiments determined that copper and zinc in any 
form, added to, or in contact with iron in any form, increased the 
corrosion of a covered iron surface 60 per cent in a given time; copper, 
without the zinc element, 40 per cent. 

The irregularity of the distribution of the atoms in a compound 
iron-oxide coating — their difference in size and character — deter- 
mine the points of corrosion, which once established end only with 
the complete failure of the coating. 

From five to ten per cent of the sulphate of lime (CaOSO,) is 
generally found in iron-oxide paints or pigments. It is usually speci- 
fied by the consumer that it shall be fully hydrated, or that it shall 
contain not more than one part of water. The effect of the moisture 
is to aid any sulphur element present to commence promptly its 
work of disintegrating the coating. From its great covering power, 
the ease of grinding and mixing it with the iron oxide, and its cheap- 
ness, one third of the weight of the pigment is frequently composed 
of this substance, particularly in the tint colors. The greater the 
amount of this sulphate of lime, the sooner the destruction of the 
coating. 

The copperas oxides of iron stand remarkably well upon wooden 



IRON^XIDE PIGMENTS, 37 

Burfaces. The brown oxides stand the best upon ferric bodies. The 
Venetian red oxide, from the old iron-oxide mines, had a peculiar 
preservative action on the surface of wood. Two or three coats 
from the pure materials have outlasted the record of their appli- 
cation and the lives of the painters that spread them. These oxides 
and white lead form a hard mastic covering, and unless spread over 
unseasoned or wet wooden surfaces, are not liable to blister or peel. 

Many of the irreconcilable discrepancies in the use of iron-oxide 
paints can be attributed to the careless method of preparing them. 
In general practice, it is never ground with the oil, and but seldom 
machine-mixed. The dry pigment in the ratio of six to seven pounds 
and about the same weight of oil (or three-fourths of a gallon) are 
placed together in a tub, and after a few hours of soaking are simply 
stirred up and spread. If any large quantity of paint is so pre- 
pared, it is almost impossible to secure thorough incorporation of 
the pigment and the oil, owing to the different specific gravities of 
the several substances composing the pigment, which vary from 
2.2 to 4.9. This manner of mixing is strongly recommended by 
the iron-oxide trade to secure its use, at the expense of the life and 
effectiveness of their product, which many times might be more 
creditable were better care taken to render it deserving. 

The longer that iron-oxide paints are ground in the full quantity 
of oil they require to form a painty the more lasting they will be, and 
this effect is equally apparent in all pigments. 

The unsatisfactory results due to careless mixing are aggravated 
by the use of large flat brushes that act as mops to carry or slap 
on a large quantity of paint, inadequately, in this way, brushed out. 
Such brushes carry air into the coating, rendering it more porous 
in drying than it otherwise would be were heavy, long-bristle, round 
brushes employed. The same objections exist where the coating 
is applied by the air-brush or spray apparatus, only in a more marked 
degree. See Chapter XXXI. 

There are many tests for the adulteration in iron-oxide pigments 
or paints of too extended detail to be entered into here. A ready 
test for the soluble sulphate of iron in an iron oxide is to warm a 
little with pure water and filter through blotting paper. Add to the 
clear solution a few drops of hydrochloric acid and a little of the 
chloride of barium (both obtainable at any drug store). If a white 
sediment forms in the solution, the sample of iron oxide should be 
rejected. 



38 IRON-OXIDE PIGMENTS 

There are many recorded instances of the excellent results attend- 
ing the use of iron-oxide pigments. Berzelius (1838) mentions that 
on houses in Sweden, painted with iron oxide three hundred years 
ago, the coating was still in fair condition, and the wood-work well 
preserved. The wooden homes and workshops in many parts of 
the United States bear testimony equally in favor of this class of 
paints, in the general good condition of the paint and structure after 
exposures, without repainting, of fifty to eighty years. But in every 
case the cause of these excellent and exceptional results was in the 
use of better materials and better methods of apphcation than is the 
present-day practice. 

Paint-trade literature frequently cites that the tin roof of Independ- 
ence Hall in Philadelphia has been protected for the past one hundred 



and thirty years with an iron-oxide paint, givinfc this as an unanswera- 
ble argument for the prntpctive character of these pipmente applied 
to metal surfaces. The facts in this case show that the plates with 
which this particular roof was laid, as well as many others at that 
period, were double-coated with pure block tin. free from lead, Jiinc, 
or antimony, more scrupulous care beinp; taken in every part of its 
manufacture to secure a reliable product, than is now practised by 
even the best of the present-day tin-plate manufacturers, whose 
products almost universally fail by the corrosion of the utin plate 



CORROSION OF TIN ROOFS 39 

on which the tin is coated. This internal corrosion casts oflf the tin, 
and no amount or kind of paint spread upon an inferior quality of 
tin-roofing metal can prevent this internal corrosion, though it may 
conceal its presence and progress, and possibly fill up some small holes 
in the early stages of the decay. 

There are scores of tin roofs covering important buildings in the 
Canadian Provinces that have been laid for nearly a century, as bright 
and uncorroded now as when first laid, never having had a coat of 
paint to protect them. 

Pure block tin is unaffected by atmospheric conditions, almost 
as much so as copper. It was only when poorly cleaned plates, 
poor tin for the coating, acid flux instead of resin for soldering, and 
careless methods of laying the roof generally, came into vogue, that 
we began to hear of the virtues and need of an iron-oxide paint to 
prevent the corrosion of a tin roof. Granted that a good quality 
of tin roofing is none the worse for a coating of paint applied a 
year or two after the roofing is laid, yet it is quite as essential 
for the future life of the roof that the paint should be also of 
good quality. 

The original mines from which the iron-oxide pigments known as 
Indian and Venetian reds were taken have long been exhausted. 
These old mine pigments required no roasting or doctoring to develop 
their color, or to correct any acid elements in them. The reputa- 
tion of these pioneer oxide pigments, like that of the "Old Dutch 
Process" white lead, has been assumed to reach and cover the advent 
of scores of substances bearing little resemblance to their progenitors, 
except in name, and even this is not exempt from the greed of some 
modem paint-compounders, as the many prefixes and trade-marks 

bear witness. 

Fig. 7 shows the protective character of an iron-oxide paint 
applied to a railway viaduct not properly cleaned from mill-scale 
before painting, and when painted was exposed to combustion gases, 
cinders, dust, and moisture. 

Prepared iron-oxide paints are often brightened by the use of 
aniline colors, but are not durable. Burning a sample of such paints 
over an alcohol lamp will destroy the aniline, and leave the iron 
oxide its natural color, exposing the cheat. The tendency of all iron- 
oxide paints is to darken \vith age, due to the natural darkening of 
the oil vehicle by age, rather than by change in the pigment. 
While nearlv all of the iron-oxide paints are adulterated, barytes, 



40 COPPERAS (SULPHATE OF IRON, QRBEN VITKIOL). 

chalk, sand silica, and all added substances lessen the covering 
power. The clays also absorb water and hasten the decay of 
the pMnt. 



Copperas Oxide. 

The chemical composition of copperas is Fe.S0,.7H,0'"0ne part 
of iron plus one part sulphuric acid, plus seven parts of water, 

C!opperas, the waste product of many manu far tu ring processes, 
is largely used to produce an oxide-of-iron pigment by roasting the 
crj'stals. Six parts of the water are driven off by a heat of 114*^., 
but one atom is still retained at 280° F. At a red heat it decom- 
poses, giving off one part of sulphurous oxide, leaving a basic ferric 
sulphate, Fe,Oj3S0j ; and , more strongly heated , it leaves a pure ferric 
oxide known as Colcothar vitriol. 

As usually roasted for an iron-oxide pigment, from three to five 
pounds of terra alba, lime, or chalk, to one of the ore, are put in the 
roasting-fumace, heated to a high heat to expel the sulphur, which 
is supposed to combine with the Hme, forming a synthetical sulphate 
of Hme (gypsum). This combination is the same as in the roasting 
of iron ore with chalk or lime to remove the sulphur, and will be 



IRON-OXIDE PIGMENTS, OCHRE. 41 

found in detail in the list of inert pigments under *' Gypsum." The 
process is a sensitive one, the color of the product being the bright- 
red color pigments, — Venetian and Indian reds. Both colors are 
due to the degree of heat employed, the length of exposure to this 
heat, the manipulation, manner and time taken to cool the mass, 
etc. 

All of the sulphur is not dissipated by the heat nor absorbed by 
the lime in its change from a carbonate to a sulphate. The lime 
changed to gypsum or left free, being in great excess of the amount 
that is allowable in any pigment, is removed to some extent by sift- 
ing, or other means, before grinding the furnace product. 

Copperas oxide requires great care in its use, either by itself or 
mixed with the dead-color iron oxides or other pigments to bring up 
their color, as the sulphur goes into the paint with the usual results. 
No amount of free adulterants or inert substances have any material 
effect in neutralizing it. 

Copperas ferric salt (a protosulphate of iron), Coquimbite, is 
found native as a hydrate, containing nine atoms of water, FejOjSSO, 
4- OHjO. It occurs in layers several feet thick in fine-grained six- 
sided pyramid crystals. Its preparation for a pigment is similar to 
that described above, and the resulting disintegrating effect in the 
paint is not measurably different. 

Yellow Ochre. 

Ochre is a hydrated oxide of iron of a strong yellow or brown- 
yellow color, generally containing less than 40 per cent of iron oxide. 
The ochres are among the oldest of pigments. Samples have been 
obtained from Pompeii in all stages of preparation from the ore to 
the mixed paint. They were used in Greece in the time of Pliny, 
and in Old Egypt. They are the most stable of the yellow colors, 
and are the principal pigment in the present freight-car colors. 

Ochres are clays tinted with the oxide of iron and manganese, 
and hygroscopic in character, carrying from five to fifteen per cent 
of water. Dried artificially to expel the water, they change color 
to pink or red the same as all other iron-oxide substances. 

Their yellow color is chiefly due to the iron oxide, and the more 
of this they contain the darker the color. The brown color is due 
to the manganese oxide. All ochres contain some amount of this 
oxide. The darker colors (unroasted) contain the most manganese, 
and are good driers for use with linseed-oil. 



42 IRON-OXIDE PIGMENTS, OCHRE. 

The covering power of the ochres depends upon the amount of 
lime or chalk in them, which reduces the coloring power by rendering 
them translucent. They require from sixty to eighty per cent of 
oil to form a paste, and the added quantity of oil to make them spread 
makes them slow driers. They blacken a little in time exposed to 
sunlight, but the change in tone is evidently more from the darkening 
of the oil than from a change in the pigment. 

The best brands of ochre are the French: 

Composed of clay 69.5 to 73.8 per cent 

Oxide of iron and manganese 23.5 " 25.6 " " 

, Water 7 0" 9.5 " " 

French ochre has a large spreading power, as it absorbs a large 
quantity of oil, and it holds well to wooden surfaces. It should 
be ground in raw linseed-oil, and if a thinner is required, raw oil 
should be used. White ochre has the property of holding well to 
wooden surfaces from the large amount of oil taken up by it, but 
does not bond well to any overlying coat of white lead, and tends to 
cast it off by peeling. This action can be avoided by using a small 
percentage of white lead in the priming ochre coat. 

The English Oxford and stone ochres are among the best brands. 
The Roman ochres grade with the best Havre, while the lower grades 
of French ochres are poor and possibly lower in covering power than 
the best Bermuda (Virginia) or other American brands. 

The name of an ochre signifies, like all other paint names, little, 
imless the material is furnished from a responsible business firm. 
Even these cheap earths have to bear a share in the general burden 
of adulteration that »s the order of the day, by an added quantity 
of clay, chalk, and barytes (the latter to give weight), but all injure 
the covering power of the ochre. Their presence is usually denoted 
by the increased amount of oil required to bring the dry ochre to a 
paste. 

There are many mixtures of ochre as the basic pigment for a 
ferric coating other than those classed as freight-car colors. One 
recommended by Dr. Dudley, and used to some extent upon the 
Pennsylvania Railroad, has decided superiority over the general 
brands of iron-oxide paints marketed under the many alluring and 
always misleading trade-names. 

Dr. Dudley's formula is: French ochre, 39 lbs.; lampblack, 1 lb.; 
japan, as drier^ 6 lbs.; raw linseed-oil, 54 lbs. (or 6f to 7 gallons. 



IRON-OXIDE PIGMENTS, VmBBR. 43 

according to time of the year that the paint is to be spread). Hot 
weather requires the least oil. 

The better brands of ochre as the basic pigment for freight-car 
colors form very durable paint coatings, whose life is generally equal 
to that of the car. The cheaper grades were formerly used to a great 
extent as cheap paints for tin roofs, but the large amount of free sand, 
lime, and other uncombined mineral substances, acids, and moisture 
that they contained, with the coarse way they were calcined and 
groimd, rendered the coatings short-Uved and unsatisfactory. They 
required a large amoimt of oil to spread them, even with a white- 
wash brush. They dried or hardened rigidly, did not bond to the 
tin, and the rate of expansion and contraction from the action of 
the sun was so materially different from the tin they covered that 
they soon cracked, blistered, and flaked off. 

Uniber. 

Umber is an argillaceous brown hematite ore, essentially 
2Fe303Si02H20, with alumina and manganic acid. Specific gravity 
2.2. Originally obtained from Umbria, Italy, now chiefly from 
Cyprus. 

As a pigment it is used in both its raw or natural state, and when 
calcined is known as burnt umber. When calcined at a low heat it 
turns a dark brown; a stronger heat dehydrates it, turning it to a 
red brown and softening it. As a ferric paint it has no special quality 
other than the ochres. The cheapness of the pigment is more than 
offset by the amount of oil it requires for a good spreading paint. 

Umber is used as a drier in boiling linseed-oil, and furnishes an 
oil of good color; but imless used in large quantities, does not make 
80 rapid a drjdng oil as the lead, zinc, or manganese driers. 

Spanish Brovm. 

Spanish brown, an iron oxide or ochre, containing thirty to fifty 
per cent of clay, is inferior in color and covering power to umber, 
but is of lasting value for a roofing paint, as the clay, which has at 
all times a strong affinity for moisture, will, when properly calcined, 
take up seventy to eighty per cent of oil, and this oil, protected from 
the sun and air, in turn protects the covered roofing metal thoroughly. 
On vertical surfaces, however, less oil must be used, else the coating 
will be liable to crawl before it is dry, and the ochre, not being so 



44 IRON-OXIDE PIGMENTS, SPANISH BROWN. 

well protected as on the horizontal surface, will absorb moisture^ 
soon corroding the ferric surface. If two or more ochre coatings 
are spread over one another, the last coating is liable to ped; hence, 
for tin or metal-covered roofs, one heavy coat is better than two or 
more thinner, unless the latter are appUed after an interval of a 
year or more. The more elastic the first coating of this pigment, 
the more durable; but the greater will be the tendency to cast off 
or crack the second or other coats. 



CHAPTER IV. 

RED LEAD. 

Metallic Lead. Symbol, Pb. Atomic Weight, 206.9. 
Specific gravity, pure 11.445; commercial 11.335 to 11.3S8. 

Lead occupies an important part in the arts and manufactures 
of the day, and requires a greater range of chemical and mechanical 
processes for its production as a pigment and more care in preparing 
and applying it for a paint than any other pigment. 

In its mineral form it is associated with all of the noble metals, 
also with copper, tin, zinc, bismuth, antimony, arsenic, etc. Some of 
the baser metals are always present in commercial pig lead, and affect 
the character of the pigments prepared from it by the processes of 
calcination, oxidation, sublimation, corrosion, and precipitation. 

There are twenty ores of the metal known to the mineralogist, 
but metallic lead is produced from the five following minerals, the 
analyses of which indicate not only the character of the ore but also 
of the metallic lead and pigments made from them: 

Sulphide of Lead, PbS (Galena, Blue-Lead Oke). 

Oxide of lead 81 .80 to 85. 13 per cent. 

Oxide of silver 0.08 " 0.02 " " 

Oxide of zinc 3.59 " 2. 18 " " 

Sulphuric acid 16.40 " 13.02 " " 

Sulphate op Lead, PbS04 (Anolesite). 

Oxide of lead 71.00 to 72.46 per cent. 

Oxide of iron 1 .00 " 0.09 " " 

Sulphuric acid 26.09 " 24.08 " " 

Water 0.51 " 2.00 " " 

Carbonate of Lead, PbCOj (Cekussite). 

Oxide of lead 66.00 to 84.76 per cent. 

Oxide of iron 2.30" 0.00 " " 

Oxide of alumina 16.30 " 0.00 " " 

Carbonic acid 13.00 " 16.49 " " 

Water 2.20" 0.00 " " 

45 



46 RED-LEAD ORES AND OXIDES. 

Phosphate op Lead (Pyromorphite). 

Lead (metallic) 7.80 to 7.39 per cent. 

Oxide of lead, 73.22 " 74.50 " " 

Phosphoric acid 16 . 76 " 15 . 94 " " 

Chlorim 2.67 " 2.54 " " 

Arsenate of Lead (Mimetesite). 

Oxide of lead 74 . 96 per cent. 

Arsenic acid 23.06 " " 

Chlorin 2.44 " " 

Metallic lead forms five oxides that in one or more forms are the 
result of the heat or chemical changes produced in the metal in con- 
verting it into a pigment. They are: 

Lead. Oxygen. 

The suboxide PbjO =96.277 per cent. 3.723 per cent 

" protoxide Pb.O «92.822 " " 7.178 " " 

" red oxide (minium)... Pb,.04 = 90. 630 " " 9.370 " " 

" sesquioxide Pb,.Oj = 89 . 606 " " 10.394 " " 

" dioxide or peroxide. ..Pb.O, -76.375 " " 23.627 



(t tt 



Red Lead, or Minium^ is the principal pigment produced from 
the oxides, its specific gravity being 8.5 to 8.94, depending upon 
the purity of the lead "Massicot'' or litharge, from which it is made. 
The color also depends upon this point, but in a greater degree upon 
the temperature employed in the oxidation of the material, the 
uniformity of the heat, the manipulation of the material in the fur- 
nace, the length of time exposed to the heat, and the rate and manner 
of cooling down the furnace and its contents. 

Special furnaces and processes are required in its preparation 
that differ materially in the various countries where red lead is pro- 
duced, also in different manufactories, and there is a difference in the 
materials employed. 

Briefly described, these processes are: First, by the cupellation 
furnace which converts metallic lead into litharge in about 24 hours. 
Second, a reverberatory furnace in which litharge is converted into 
red lead. Third, a reverberatory furnace or oven that reduces metallic 
lead into litharge. In any of these processes the litharge is the first 
product, and is always formed whenever metallic lead is heated to 
about 900*^ F. for about 24 hours and freely exposed to a current 
of air, the material being continuously stirred during the heating 
process. This crude product or litharge is a coarse granular substance 
that upon further heating is fused into a crystalline mass. When cold 



BED-LEAD MANUFACTURE. 47 

this mass when broken up is in the form of thin yellow or brown 
scales, and in this state is known as flake or glass-makers' litharge. 
This, when ground in water and dried, changes its color to a buff, and 
is the ordinary commercial litharge. 

When the crude litharge powder is again moderately heated in a 
reverberatory furnace or oven and exposed to a current of air with 
continuous stirring for from 26 to 48 hours, or until a sample drawn 
at a low red heat appears of a dark-red color turning to a bright red 
on cooling, the furnace is closed and allowed to cool slowly; a con- 
dition most essential to success in the color, that if not satisfactory, 
requires a reheating and cooling of the product now known as red lead. 

Red lead made from Htharge (from the imperfect oxidation of 
the litharge) contains a larger amount of the protoxide of lead than 
that made from the carbonate or white lead, where, on account of 
the finer condition of the material, the oxidation is more complete, 
more quickly effected, and generally of a better color and quality. 

With this complex chain of operations there are many trade 
secrets to secure not only a uniform quality of the pigment, but its 
color. The latter nearly always remains an uncertain element even 
with the best of attention given during the whole process. 

Red lead is found native in many localities mixed with the other 
ores of lead, probably resulting from their oxidation by natural causes. 
Chromate of lead (Pb.Cr04. Specific gravity, 4.6 to 5.2), the neutral, 
or meta-chromate, known as crocoisite or lehmanite, is a native red-lead 
ore found in commercial quantities in many parts of the world. It is 
in the form of translucent crystals of a yellow color with various 
shades of other colors, and is associated with decomposed gneiss and 
granite. The method of converting these red-lead ores into pigments 
need not be described here. All of the yellow and red chromates of 
lead are obtained from crocoisite. They are strong colors, and do 
not decompose on exposure to the air or light. 

Red lead is one of the heaviest and most expensive pigments, 
also the most difficult to prepare for a paint or to spread. It is more 
susceptible to adulteration, and is more adulterated by interests inimi- 
cal to its reputation, than any other pigment, with the possible excep- 
tion of its sister-product, white lead. There are many well-authen- 
ticated instances of its perfect protection of important structures, 
and a great number of its failure in locations where the faUure can 
be directly traceable to causes detrimental to the success of any 
pigment. It has been, and is, condemned for causes directly trace- 



48 RED-LEAD ADULTERATIONS. 

able to its improper preparation and application, and where the 
failure should have been foreseen by the engineer or master painter, 
and where carelessness, indifference, or ignorance of the conditions 
to which the coating was to be subjected, were the prime factors of 
its non-success. 

When obtained from a reputable manufacturer and properly 
prepared with a suitable vehicle and spread under known condi- 
tions of its future service, it has proved to be one of the most reliable 
pigments. Its color is distinctive, hence it is not favorable to the 
use of adulterants. Brick-dust, iron oxide, and barytes tone down 
its high color, but are detrimental in all other respects. Chalk, 
g\''psum, and other light-color and low-specific-gravity substances 
are often added to correct the tendency of red-lead paint to "set" 
in the paint-pot while applying it, or to prevent its "creep" when 
spread upon vertical or inclined surfaces. All such adulterants are 
easily detected ; they do not prevent the set or crawl of the paint, 
and are the principal cause of the failure of the coating. For the 
foundation coat upon ferric bodies, it will cover about as much 
surface as any other paint applied under the same conditions and 
with the same effort on the part of the painter to brush it out. The 
latter factor is frequently too small to ensure success with even a 
whitewash. 

As a first coat on ferric bodies, applied at the w^orkshop, its color 
shows at once any material injury to the coating due to the usual 
handling and transportation, also readily indicates if the grease and 
dirt due to the machining processes in the shop or received during 
transportation and erection have been properly attended to or not. 
It is a prime watch-dog in this respect. 

Its tendency to settle in the paint-pot, also "to set," and the 
necessity for constantly stirring it up by the painter, probably lessens 
by a small amount the number of square feet of surface he can spread 
in a day. This is probably an objection to its use, but is offset by 
the many points in its favor. The rapid set of red lead when mixed — 
a peculiarity of this pigment — is as objectionable in a paint before 
it is applied to the surface as the set or hardening of hydraulic cement 
on the mortar-board. When either the paint or mortar has set before 
being spread, it is useless for its intended purpose. If the set of either 
is broken up by stirring and they are then applied, both may appear 
to be of the same nature as before setting, but they are injured beyond 
recovery. The mortar will not be much better than a wetted sand, 



SETTING OF RED LEAD. 49 

and will not again bond the sand or to the masonry. The red lead 
will not recover its combining power that ensures the mutual bond 
between the atoms of the pigment and to the surface covered, be it 
wood or metal. When in this condition, if it is applied to a metallic 
surface, it "crawls," as it is called, and presents an appearance more 
like that of curdled milk than a paint, and the actual protection of 
the body covered is due to the vehicle only. It will add but little 
to the covering power other than what any adulterating substance 
would do. 

The setting of red lead is due to two chemical reactions, namely, 
a combination between the litharge of the red lead and the glycerine 
element in the oil ; also a combination between the fatty acids of the 
oil and the litharge, forming a lead soap, quite a firm substance, but 
one not favorable to the durability of the paint. 

Many of the failures of red-lead coatings, if rigidly traced to their 
source, would no doubt be found to have been caused by the care- 
lessness of the painter in not keeping the paint well stirred during 
its application, or in preparing too large a quantity for immediate use, 
or by using the paint left over from day to day, or from another job. 

Red-lead and Lampblack Mixtures to Delay "Setting.*' 

Iron oxide, zinc oxide, barytes, gypsum, etc., added to red lead to 
prevent "setting," are objectionable and ineffective, as before noted. 
Lampblack from J to 1 ounce per pound of red lead delays the set- 
ting action and enables the red lead to be prepared as a paste to be 
used in the immediate future, when it is thinned by additional oil at 
the time and place of using. The bright red of the minium is modi- 
fied by the lampblack to a chocolate color that may be light or dark 
according to the quantity of the lampblack used. Lampblack of 
itself is an excellent pigment, is electrically passive or neutral to all 
pigments, and by natural formation is so finely divided that it mixes 
easily and thoroughly with the oil and red lead without deteriorating 
the quality of either. 

Many painters have reported a difficulty with the use of red lead 
and lampblack: that with 1 ounce of lampblack per pound of red 
lead the paint would not dry promptly for shop work without the 
use of japan or turpentine driers. Bridge engineers, however, report 
that red-lead paints carrjdng 10 ounces of lampblack to 12 pounds 
of red lead have stood for three or four months without setting or 



50 RED-LEAD AND LAMPBLACK MIXTURES 

hardening by simply agitating or rolling the barrels daily. In one 
case the paint barrels were left undisturbed for four months^ and 
though the red lead had settled in some degree, when it was stirred 
again the paint spread as well and dried as firmly as though freshly 
mixed. 

Lampblack containing sulphur in any appreciable amount should 
never be mixed with red lead for a coating. Ground soot from chim- 
neys or furnace-flues, ground bituminous coal or coke, or the soot from 
most of the petroleum or heavy mineral oils, contain sulphur enough 
to cause the prompt failure of a red-lead coating, even when used to 
the amount of i of an ounce per pound of red lead. 

Mr. Ball, master painter Pennsylvania Railroad, 1897, reported 
the result of some of his experiments with "protective paints for 
metallic parts of cars and trucks," viz. : 

First. Red lead and raw linseed-oil, with litharge as a drier. 
Second. Red lead and raw linseed-oil, no drier. 
Third. Red lead and lampblack, equal parts. 
Fourth. Red lead one part, lampblack three parts. 
Fifth. Mineral brown (red oxide of iron;. 
Sixth. Mexican graphite. 

All paints were mixed from the same quality of raw linseed-oil; 
none but the first had any drier. 

At the end of fifteen months of atmospheric exposure at the car 
shops their condition was as follows: 

Number One had failed in two or three spots. 

" Two was intact and appeared to be in condition to resist 

a number of years' wear. 
" Three had scaled off in a number of places, showing rust. 
" Four was in a still worse condition. 
" Five was completely gone. 

" Six was in perfect condition, the same as the straight red- 
lead sample. 

The slow setting and drying mixtures of red lead and lampblack 
are probably better for field work, where slow dr>dng is of little moment, 
unless bad weather conditions are to be met, than for use at the shops 
where the transportation requirements govern. Thinning the red- 
lead and lampblack paste wnth boiled oil when the paint is appUed 
will add the necessary drying element to the paint without the use of 
turpentine. 



RED-LEAD AND LAMPBLACK MIXTURES. 51 

Spirits of turpentine is objectionable in all paints, as its evapo- 
ration leaves the coating more porous than it would be if the paint 
dried naturally; and it deadens the gloss. Japan driers for red-lead 
paints are less objectionable than the turpentine. They are heavier 
and add resinous matter to the paint in drying that is of the same 
character as the vehicle with which the red lead is ground. The 
drying of japan b principally by resinification and not wholly by 
evaporation, as b the case with turpentine; with the added japan 
the period of drying can be governed as required by using a small 
proportion of raw linseed-oil. 

Excerpts from a trade catalogue * relative to the use of turpen- 



tine and red lead say: "We are not prepared to advise the use of 
turpentine in shipyards, where, owing to time contracts, it is often 
necessary to paint in damp or freezing weather, though the practice 
in the Marine Department of the Maryland Steel Co. from many 
years' experience in the use of red lead is, viz.: Use three parts of 
linseed-oil (raw, presumably) and one part of turpentine for the first 
and second coats, with sufficient drier (presumably japan) to set 
well in twenty-four hours, allowing five to six days between the 
coats. The third coat to be an oil coat without drier." 

* "How to Use Hed Lead." National Lead Company. 



52 RED-LEAD PAINT MIXTURES, 

"Theoretically, and for dry, bright weather, red lead should 
be used with raw linseed-oil and no drier, red lead itself being a natural 
drier. By so doing, the chemical union between the pigment and 
oil is most complete and the resultant paint is more durable. How- 
ever, for ordinary service add a small quantity of japan or use boiled 
linseed-oU. By so doing a more viscous vehicle is had, which better 
sustains the heavy particles of the red lead, thereby preventing its 
running on vertical surfaces, and possibly givmg greater covering." 

Red-lead Paint Mixtures. 

Many raflway engineers favor the following mixture of red-lead 
paint for their structures where the time of drying is of little moment. 
For the first or priming coat, 20 poimds of red lead and 1 gallon of raw 
linseed-oil. No driers. For the second and third coats: a paste 
made from 60 pounds of hydrated sulphate of lime, 30 pounds of lamp- 
black, 5 pounds of red lead, making 100 pounds of pigments, to which 20 
gallons of boiled linseed-oil are added, making 30 gallons of paint. 
This makes a fair drying paint of a dark or dirty grayish-brown color, 
weighing about 8J pounds per gallon. All of the power in this paint 
to prevent corrosion lies in the red lead and lampblack, the sul- 
phate of lime adds the stuffing for quantity without contributing 
anything of a protective character to the mixture, and the covering 
power is poor. 

If any of the low-carbon amorphous graphites were substituted for 
the sulphate of lime in this formula, the paint would be better in 
all respects for coating ferric bodies. 

Mulder's Experiments vxith Cheap Red-lead Paints. 

Mulder, in his experiments to produce a cheap paint, used boiled 
llnseed-oil containing 2J per cent of red lead. He used 100 parts 
of this oil with every one of the following mixtures. The iron oxide 
used was Cartier's (Belgium), that analyzed as follows: 

Iron oxide 68.27 per cent = 47.79 per cent of metallic iron. 

aay 27.00 



Marl 0.27 

Chalk 0.40 

Water 2.75 

Undetermined. ... 1 . 31 

100.00 



a 
ti 
it 
tt 

tt 



27 . 67 " " mineral substances. 



CHANGES IN RED-LEAD PAINT, ACTION OF SULPHUR. 53 

200 parts of pulverized fine sand, 5 parts red lead, gave a paint 
of some merit. 

Red lead 25 per cent, iron oxide 40 per cent, gave a very good 
coating. 

20, 40. 60, part-s of red lead to 100 parts of iron oxide gave excellent 
results. 20 to 90 parts of red lead to 50 parts of pulverized red 
roofing tiles 'gave a thick heavy coating. 

40 parts of red lead and 100 parts of pulverized red rooflng tiles 
gave an excellent coating. 

20 to 90 parts of red lead and 100 parts of pulverized ironstone 
gave a paint of distinguished excellence. 



Fio. 9.- 

Other mixtures of red lead favorably reported upon by experi- 
menters are red lead, zinc oxide, and Blanc-FIxe. The latter sub- 
stance serves to hold up the red lead in the mixture during the first 
stage of drying and prevents its "creep." Amorpho\is graphite 
Is also used instead of the Blanc-Fixe for the same purpose. None 
of these substances or any others employed for this purpose can be 
so mixed that they will not be subject to the influences that destroy 
all compounded paints, as mentioned elsewhere in this work. (See 
Chapters XXVII and XXXII.) 



54 CHANGES IN RED-LEAD AND ZINC-OXIDE MIXTURES. 

The bright lustre of red lead is often toned down by Venetian 
red. This pigment, if it could be obtained from a natural ore, is a 
very desirable one, but the mines that furnished it in former days 
have long been exhausted as a commercial source of supply, and 
the copperas reds have taken its place. These contain a large amount 
of sulphuric acid loosely held in combination (see Analysis, Chapter 
III), and their use with red lead is as disastrous to the paint as the 
direct effect of hydric sulphide. 

The effect of the sulphur element in both cases is greatly influenced 
by heat, a few weeks' exposure to hot summer temperatures being 
all that is necessary to destroy the coating by rendering it brittle 
and easily removed by the hand, or else causing it to peel in strips. 
This action of the sulphur is sometimes said not to be due to a change 
for the worse in the red-lead atoms, but to the change in the ve- 
hicle. But the effect of sulphur upon dried linseed-oil is almost 
nUf and the oil-film is always nearly transparent of itself, whatever 
pigment may be associated with it. Any change in the pigment 
is denoted by a change of color in the paint, whatever may be the 
cause of the change. The change in the pigment simply shows 
through the thin film of dried oil quite as reatlily as though it were 
of glass, and generally indicates a speedy dissolution of the coating. 
In many mills and workshops all of these conditions — heat, sulphurous 
fumes, and saturated atmospheric elements — are present, and red-lead 
coatings in such locations are short-lived. 

The sulphur element, whether in the oil or the red lead, or from 
any other source, renders the paint liable to dry on the surface only, 
and the inner portion of the oil that encloses the pigment-atoms 
remains soft and, therefore, more sensitive to any destructive influ- 
ences that reach the coating. 

From 28 to 30 pounds of red lead to a gallon of oil are necessary 
to make a good red-lead paint, for even when well ground it is liable 
to streak, curdle, or nm, and is difficult for the painters to spread. 
The bulk of tlie red lead is so small compared \^ith an equal weight 
of any other pigment per unit of covered surface, that the atoms of 
the red lead are well housed in the oil and better protected. Hence, 
when the conditions are favorable for a red-lead coating, it proves to 
be a more durable one than coatings made from other pigments 
that carry (as many of them do) in their composition the elements 
for their dissolution. 

Fig. 10 illustrates the character of red-lead coatings when not well 



CHANGES IN BED-LEAD AND ZINC-OXIDE MIXTURES. 55 

worked or brushed out in spreading. The porous character would 
disappear when spread on any surface other than glass. 

The United States and other Governments have favored in the 
past a mixture of two or three parts of red lead and one of zinc oxide 
for the protective covering for lighthouses and seacoast Iron struc- 
tures. These coatmga are harder than red lead alone, and better 
lesist the action of salt-water epray, fog, and the abrasion from the 
sand-blast usual in such locations. But the result of some thirty 



traoBpartnt in spots, although to the eye It looks solid. (M. TcNih.) 
years' experience with these mixtures has led to their abandonment 
for the reason that the oxide of zinc in the coating changed to a car- 
bonate of zinc, and by its increase of volume disrupted the dried coat- 
ing, exposed the ironwork, and the increase in the corrosion was 
markedly greater than Tvith red lead alone, or red lead and silica, 
or red lead and graphite coatings. 

Red-lead coatings soften metallic tin, hence for tin roofs they 
have not proven so durable as iron-oxide or graphite coatings, and 
are too expensive for that purpose. The effect of the red lead for 
such purposes is to form a white oxi<le of tin by the galvanic action 
between tho two metals. The oxide of tin is free, and having no 
vehicle incorporated with it, Ls easily washed out by storms, leaving 
the iron plate entirely unprotected. 



66 RED-LBAD MIXTURES. 

Fig. 11 illustrates the action of a red-lead-compound paint showing 
the character of that class of coatings for the protecting of surfaces, 
especially a ferric one. 

Red-lead compositions are extensively advertised to keep indefi- 
nitely without setting, and that are ready for use at any time without 
further mixing or preparation. In all such mixtures, the red lead. 



Fig. 11. — Photomicrograph X 100 of a film of dried paint taken from an iron 
pillar siiowing rust blisters. The dark spot is red lead and a fissure 
runs through the centre. The zinc oxide and white lead are white and are 
intact. (H. Toch.) 

or the oil, or both, are adulterated and will be found to be compara^ 
tively short-lived and unreliable whatever may be the guarantee, 
which in general lays more stress upon the extraordinary large surface 
that can be covered than the permanent character of the coating. 
As well expect a hydraulic cement ready mixeil to be a suitable article 
for engineering use as an " alwaj^-ready " red-lead paint. 

The setting or soliilification of a pure or nominally pure red-lead 
paint is a characteristic chemical union between the oil and the lead, 
and without this action the paint is worthless. This chemical action 
is sought to be simulated in all compound paints by the liberal use 
of driers either incorporated in the vehicle by heat or by being 
introduced through the bung-hole of the barrel. The setting must 
take place eventually, and the better paint will be the one in which 
it is definitely provided for, and not left to the haphazard operations 
around the bung-hole. 



RED-LEAD MIXTURES READY FOR USE. 57 

One gill of crude mineral oil or heavy refined petroleum added 
to a gallon of red-lead paint will delay the setting of it mdefinitely. 
It will dry superficially, as the oxidizing power of the red lead will 
ensure that essential, but the petroleum will always remain viscid in 
the coating and eventually destroy it by peeling soon after an exposure 
to a strong sunlight or heat, following or followed by a lower damp 
temperature or a storm. 

Red lead, either in the form of a pigment or paste, when quoted 
as being "second quality," can be regarded not only with suspicion, 
but with a certainty that it is greatly adulterated or poorly oxidized 
from impure lead, or not properly washed or pulverized. 

First-class manufacturers of red lead have no second-quality 
product that they are willing to have bear their brand or seal. There 
are a large number of red-lead corroders in the United States, and 
to the author's knowledge only one of the number advertises a second- 
class product. It may also be of interest to note that the United 
States Bureau of Construction, in its orders for red lead, specifies 
the make of one corroder as the standard of quality to which all 
tenders must conform. 

A good red lead as it comes from the manufacturer is finely pul- 
verized, as this point in a great measure governs the setting and 
running (creep, crawl, or curdling the painters call it). The atoms 
should be opaque, which indicates a good covering or light dispersing 
power. If the atoms are crystalline and more or less trans- 
lucent, "the paint will have a tendency to "tack." This effect does 
not always indicate that the pigment is deficient in other respects to 
form a durable coating; for the "tack" is sometimes due to the quality 
of the oil, and that the red-lead manufacturer has seldom anything 
to do with. 

The adulteration of red lead and litharge can be readily ascer- 
tained by digesting a sample in a warm solution of nitric acid; the 
adulterants will remain undissolved. 

BoiUng hydrochloric acid will extract the iron oxide from the 
residue. If adulterated red lead is ignited there remains a mixture of 
yellow lead oxide and the red or other colored substances that have 
been added to the red lead. 

Red lead boiled in hydrochloric acid is slowly converted into the 
chloride of lead with an evolution of chlorine gas. Dilute nitric 
acid only slowly dissolves red lead, leaving a brown powder. 



5S RED-LEAD ADULTERATIONS AND TESTS, 

Salt creates a chemioal action on red lead that is liable to blister 
the coating and reduce the red lead to a metallic state. 

Grimshaw recommends a mixture of red lead with painters' 
sizing to cover pine knots or yellow pine woodwork, instead of the 
usual shellac varnish. It forms a heavier coating than shellac, is 
equally or mpre resistant to the pitch, and is less liable to blister. 

A gallon of pure linseed-oil will require not less than 20 pounds 
as a minimum quantity of pure red lead to 30 pounds as a maximum 
quantity for a reliable red-lead paint which will cover from 750 to 
1200 square feet of metallic surface. These quantities of material 
at once remove red-lead paint from any comparison of cost with 
the oxide-of-iron and many mixed paints — principally in the form 
of proprietary goods, the ingredients of which are only known to the 
makers, and the character and performance of which will vary in 
quite as erratic a manner as the price pairl for them. 

The protective qualities of a well-oxidized pure red-lead and a 
pure oil paint, properly applied to any structiu-e under any exposure, 
except to the action of hydric-sulphide gas, cannot be gainsaid. But 
what effect other than failure of it can be expected when a govern- 
ment engineer in charge of an important hydraulic construction, 
after cleaning the metal part of the work by the sand-blast, coated 
it with the following paint? "Red lead, 40 pounds, mixed with three 
finis of water to one gallon of raw linseed-oil for the first coat, and 
for the second coating, red lead, 40 pounds, three pints of water, 
three ounces of lampblack mixed with enough turpentine to make 
a paste, and one gallon of raw linseed-oil. It was found necessary 
to first moisten the red lead with water to prevent the paint from 
streaking and sagging. Without the water, a large proportion of 
turpentine and drier would have been necessar}'', and this was con- 
sidered injurious to the life of the paint. In warm weather a slightly 
less quantity of red lead could be used" (or more water?). 

Many kindred examples of such "how not to protect" structures 
can be cited, none, however, more conspicuous than the above when 
the engineer in charge and the character of the work are considered. 

The substitution of water for turpentine in the amount here 
noted in order to prolong the life of the paint will be welcome news 
to the many manufacturers of that class of patent or proprietary 
paints who have heretofore deemed an addition of 20 to 25 per cent 
of water to the 7 to 8 per cent present in their green linseed-oil as 
about the limit of a whipped-in-oil vehicle. They can now proceed 



LITHARGE, QUALITIES AND ANALYSIS. 59 

to water their stock of paint to a point where even the water in the 
financial part of their enterprise will seem in comparison but an 
insignificant pool. (The effects of watered oil is further considered 
in Chapter XXV.) 

Litharge f PbO (Protoxide of lead). Specific gravity, 8.50 to 9.00. 

Litharge as the first product in the oxidation of metallic lead 
to form red lead has been described. Another source of litharge is 
from the scum of melted lead or that from the smelting of silver- 
bearing ore.s. It is formed as an oxide by exposing to the roasting 
heat of a furnace the slag, or "matte," that on cooling forms into 
white or flake litharge. The part that hardens last is called "Massi- 
cot" or "levigated" litharge, and is ground in water, dried, and 
made ready for the market. It is a yellowish-red substance or an 
amorphous powder, and crystallizes in fine six-sided scales or plates. 
It is a yellow or reddish protoxide of lead, partially fused and semi- 
transparent. The yellow is the fused or hard pieces that require to 
be ground and levigated; the red atoms are the flakes. The difference 
in the color arises from the mechanical condition resulting from the 
manner and difference in cooling the roasted product, a rapid cooling 
giving a yellowish color, a slow cooling a reddish one. 

By analysis it consists of the protoxide of lead, (PbO), 

94.68% to96.20%='89.28% to 87.86% of metaUic lead. 
6.93% " 6.82% of oxygen. 

2.89% " traces " the oxides of iron, zinc, copper, antimony, bismuth, etc. 
4.36% " traces " arsenious, silicic, and carbonic acids. 
0.49% " traces " lime. 

The special furnaces employed and the manipulations of the 
charge during the heating and cooling processes applicable to the 
manufacture of red lead are requisite for a reliable litharge; also 
the same care in grinding and the subsequent operations to prepare 
it for a pigment or for other uses. Its integrity when used in a paint 
is affected by the same causes that affect a red-lead coating. It is 
adulterated, if possible, in a more barefaced manner than any red 
load. 

Orange mineral is made from the litharge "Massicot," also from 
white lead. In many cases refuse white lead is used as the base 
material. The material is placed in a reverberatory fimiace and 
exposed to a moderate heat and a current of air and stirring as usual 
for producing red lead. 



60 ORANGE MINERAL. I 

The carbonic acid in the white lead is expelled, leaving a protoxide i 

of lead which absorbs more oxygen and produces a red lead of a i 

lighter color than that made from litharge by reason that the oxida- 
tion is more complete. 

Paris red is prepared by roasting the carbonate of lead to a litharge, 
the difference between the Orange mineral and Paris-red pigments 
being that the latter retains a little carbonic acid in its composition 
due to the different degree of heat employed in the furnace and the 
manner of cooling the product. Vermilionette is an orange-red 
pigment formed from the oxide of lead. 



CHAPTER V. 

WHITE LEAD. 

"White Lead, FbGO| (Hydrated Carbonate of Lead). Specific gravity, 

6.465 to 6.480. 

The native anhydrous meta-carbonate of lead, (PbCOj), called 
white-lead ore or eerussite, when pure is found in colorless crystals 
of the trimetric system. It is found in commercial quantities in all 
parts of the world where mineral lead ores are mined for smelting pur- 
poses. Pliny mentions the use of a native ceruse fo\md on the lands 
of Theodotus at Smyrna. 

The proto-sulphide of lead, (PbS), is the blue-lead ore (Galena) and 
is the principal source for the supply of metallic lead. White lead 
and the red oxide of lead are next to the oxide of iron, ochre, um- 
ber, and sienna, the oldest-known pigments. Dioscorides (b.c. 400) 
Pliny, and Vetruvius all mention the production of white lead by 
exposing metallic lead to the vapor of vinegar, giving the product the 
name of "Cerusa" and "Cerosa." Bergman in 1775 localized it as 
a carbonate of lead instead of an acetate, as it had before been con* 
sidered. 

White lead was used by the Egyptians as a cosmetic long before 
its employment for a pigment. 

The mining and smelting of lead ore to produce metallic lead were 
practised by the Chinese 2000 years B.C. In the smelting of lead 
ore large quantities of the lead are oxidized to the red-lead "minium," 
the use of which as a pigment antedates the knowledge of producing 
white lead by corrosion. Moses commanded the Israelites to purify 
lead (called opheret) by fire. 

The principal amount of white lead is produced by the so-called 
"Old Dutch Process." This process did not originate in Holland, 
where the recorded establishment of it does not appear before the 
sixteenth century. It was probably introduced into Holland by the 
Saracens. Venetian lead was early known for its purity and com- 
manded a higher price than the Dutch-manufactured lead on this 

61 



62 OLD DUTCH PROCESS WHITE LEAD. 

account. The establishment of the white-lead industry in England 
was almost synonymous with that of Holland, and evidently was 
introduced by Hollanders, hence the name "Dutch Process Lead." 
In this process thin perforated sheets of lead are exposed in gallipots 



Fia. 12,— Sheet-lead buckles and pot. 

contiuning a weak solution of acetic acid (water with 2} parts of strong 
acid) or common cider vinegar. The pots are placed in long tiers, 
each tier being loosely covered with boards and stacked in large num- 
bers, 9000 to 10,000 pots containing 60 or more tons of metallic lead. 
The bed of pots is then embedded in tan-bark, sawdust, stable litter, 
etc., that ferments and soon raises the temperature of the mass to 
140° or 165° F. A quantity of vinegar containing 50 pounds of strong 
acid converts 2 to 2^ tons of lead into the carbonate of lead in about 
100 days. The only attention the betl requires during the process of 
corrosion is to control the temperature of the mass by regulating the 
admission of the air to the interior of the beds by opening or closing 
the apertures left for that purpose. The corrosion is practically 
completed at the end of 60 days, but the lead is of light specific gravity, 
so it is the practice to allow the beds to remain unbroken for 30 or 40 
days more, in which time the lead acquires a proper density. If the 
lead is allowed to remiun in the beds too long, say 5 or 6 months, it 
is liable to become crj-stalline and transparent and will be of poor 
covering power. Care is necessary in the use of stable Utter or that 
from flesh-eating animals, as they are liable to change the white car- 
bonate of lead as it forms into a dark eiilphide of lead from the sul- 
phurous hydrogen evolved in the decomposition of the manure. 

At the time of stacking the air in the bed contauis about 20 parts 
of oxygen ; after 2 weeks it will contain only 17 parts ; in 5 to 6 weeks, 
7 to 15 per cent, while the carbonic-acid element will have increased 
from f of 1 per cent to 23 or 27 per cent during the process of cor- 
rosion. From 30 to 40 per cent of the lead remains unchanged, which 



OLD DUTCH PROCESS WHITE LEAD. 63 

is separated from the carbonate by passing the contents of the pots 
through a series of rolls, beaters, and screens. The corroded lead is 
then mixed with water and ground between buhr-stones to an impal- 
pable powder. Generally this part of the process is omitted by the 
quick-process lead manufacturers, because of the fine state of divi- 
sion to which it is necessary to reduce the metallic lead for these proc- 
esses. The uncorroded particles are so intimately associated with 
the carbonate that they are indifferently eliminated in the separator, 
and if run through the water-stones, will cover the face of the stones 
with a coating of metallic lead that soon impairs their grinding power 
and imparts a dark color to the product. 

If the preliminary washing before grinding is not thoroughly done 
to free it from the acetic acid (which is a drier) the powdered carbon- 
ate will dry in grains and lumps, and it may contain partly corroded 
or pure-lead particles, in which case the corrosion of them will pro- 
ceed in the paint coating from the carbonic acid in the atmosphere. 
Added adulterants of any nature cannot prevent this secondary cor- 
rosion. 

Silver in the metallic lead produces a pinkish cast in the corroded 
lead, while bismuth inclines it to a dark or gray color. Antimony, 
arsenic, iron, zinc, and other metals also have a great effect on the 
color of the corroded lead. 

After grinding, the mixed carbonate and water is mechanically 
floated to remove any coarse particles, then pumped into large settling- 
tanks ^ where it is double- washed with pure soft water and bicarbonate 
of soda in solution to neutralize any trace of the acetic acid that may 
be present. After giving the lead time to settle in these tanks, the 
water is drawn off, and the pulp lead, carrying about 24 per cent by 
weight of water, is pumped to large shallow copper drying-pans and 
the water evaporated. This drying process requires from 6 to 8 days, 
the temperature of the drying-rooms being kept at from 140° to 160° 
F. The lead product when it leaves the drying-pans is pulverized 
and marketed as dry white lead or ground in buhr-stones with 
linseed-oil for a paste or mixed paint. 

A modification of the Dutch Process Lead, known as "Pulp Proc- 
ess Lead," consists of taking the pulp lead from the settling-tanks 
and placing it in a tank of linseed-oil and subjecting the mixture to a 
high-speed mechanical stirring for a number of hours. Some of the 
water in the pulp lead is expelled, and rises to the top of the mixture 
and is drawn off; but a great part of the 24 per cent of water in the 



64 OLD DUTCH PROCESS WHITE LEAD. 

pulp is mechanically whipped into an emulsion or forced combina- 
tion with the lead. Pulp lead is decidedly inferior; even if subse- 
quently ground it does not bring it up to the standard grade of a 
white-lead product. Pulp leads are inclined to chalk more than the 
same lead submitted to the full process of drying, chasing, and grind- 
ing. They are more uncertain in taking tints, and, when applied in 
frosty weather or on exposed situations, are prone to peel. They also 
require more driers to aid in driving off the surplus water. Their low 
price is all that gives them a market. 

By the "Old Dutch Process'' the lead is neither oxidized nor 
carbonated at the expense of the acetic acid. The oxygen is de- 
rived from the air, and the carbonic acid from the tan-bark or other 
fermenting source. The vapor from the acid element as it is evap- 
orated by the heat of fermentation merely serves to dissolve the 
oxide of lead as it forms, converting it into a basic acetate, which 
is again decomposed by the carbonic acid, the acetate being thereby 
set free to act upon another portion of the lead. 

This is shown to be the mode of action by a modern process of 
corrosion, in which the protoxide of lead (PbO) is moistened with 
water containing about 1 per cent of the neutral acetate of lead 
(sugar of lead, PbO.C4Hg03) and a current of carbonic-acid gas passed 
over it, the litharge being quickly converted into an excellent white 
lead. 

There are many modifications of the "Old Dutch Process" that 
are referred to hereafter; all intended to improve the product, shorten 
the period of corrosion, and avoid the deleterious effect of the gases 
evolved from the corroding lead, upon the workmen. In fact, all 
the operations connected vnih the manufacture and use of lead 
products by any process, from the lead ore to a pigment, are exceed- 
ingly detrimental to the health of all persons so engaged, even with 
the best^known precautions. 

When honestly and thoroughly done through the long chain of 
operations called the "Old Dutch Process," the product is as fine, 
smooth, and homogeneous in character as any known pigment, and 
can be used with little or no waste. It is particularly free from 
sulphur compounds, which invariably change lead from a carbonate 
to a sulphide, to the detriment of the color and life of the paint. 

Even with the* best of care in the corrosion, "Old Dutch Process" 
lead differs greatly in character. The product from the centre of 



OLD DUTCH PROCESS AND COMMERCIAL WHITE LEAD. 65 

the stack may differ from that at the side walls, where more moisture 
is present. An excess of moisture gives the grains a sugary appear- 
ance. The evenness of temperature in the stack, due to many causes, 
also the time of the year that the corrosion is effected, governs the 
quantity and quality of the product. 

Commercial white lead is often inadequately corroded, washed, 
and ground. Pieces of uncorroded lead, tan-bark, and other sub- 
stances from the ferment-packing, are incorporated in the pigment, 
subsequent decomposition of which in the mixed paste or paint 
discolors it and shortens the life of the paint. To such an extent 
Ls the careless corrosion of lead practised, that the brand of *'01d 
Dutch Process" is called in derision **the happy-go-lucky process" 
by the advocates of the so-called "quick process." It is no longer 
a criterion of perfection in manufacture and purity, unless obtained 
directly from reputable manufacturers or dealers. Nevertheless, 
white-lead paste has held its former excellence and in some cases 
has been improved. It is also safe to say that but few corroders 
continued to use the Old Dutch Process, however desirable it may 
be to have the reputation for so doing, owing to the confidence that 
painters and users of paints have in its merits, which have been 
known and established for centuries. 

Some of the modern processes of corrosion consist in the reduc- 
tion of metallic lead into ribbons or wires, or subjecting the molten 
lead to the action of an air or steam blast, by which it is riven into 
small particles, greatly increasing the surface exposed to the action 
of the acetic acid, thus expediting the formation of the acetate of 
lead, which is afterwards corroded by passing carbonic-acid gas 
over it; the latter being generated in special apparatus, or by the 
hot products of combustion from gaseous fuel. These processes, 
while they cheapen and hasten the corrosion, have not improved 
the quality of the product or lessened the dangers to the employes. 

The product by these latter-day processes is not only coarser 
in composition, but in some cases is decidedly crystalline, and carries 
the water of crystallization, that is afterwards set free in the grinding 
process, into the paste or paint, to their detriment, and the surface 
they cover, whether it is of wood or metal. 

All gaseous fuel, unless purified by special processes, and with 
great care, as in the practice of purification of illuminating coal- 
gas, contains sulphurous-acid vapor in its composition. However 
little of this substance is taken up by the lead exposed to its 



66 QUICK-PROCESS WHITE LEADS, 

action, its detrimental effects will be seen sooner or later in the 
product. 

The sulphite of lead (PbSOj), prepared by passing sulphurous 
oxide into a solution of neutral plumbic acetate (sugar of lead), 
is a white insoluble, anhydrous powder called "precipitated white 
lead." It is of comparatively recent use and is favorably reported 
upon for a pigment. 

The hydrated carbonates of lead, formed by the direct action of 
carbonic acid on the hydrate of lead (Pb.(0H)2), differs from the pre- 
cipitated carbonates in being amorphous and perfectly opaque; 
whereas the precipitated carbonate is an aggregate of minute trans- 
parent crystalline grains. Hence the former are the best pigments; 
their greater opacity gives what the painters call "body." 

In the German, Austrian, or Chamber processes of corrosion, the 
lead is used in sheets 1"X8"X12", 1800 or 2000 sheets in a box, 8 
boxes to a chamber that may contain 12 to 24 tons of lead. The 
walls of the chamber are lined with metal and heated by steam. 
The carbonic-acid gas is made by the fermentation of vinegar, yeast, 
and other substances, ammonia, phosphate of magnesia, etc., being 
added to hasten the fermentation. Carbonic acid from burning 
charcoal and other methods are employed for generating the car- 
bonic-acid gas in great volume, for a quick corroding vapor to fill 
the chamber. 

The Kremser white or Klangenfiui), a German corrosion process, 
uses the vinegar from dried grapes as an excitant to corrosion. The 
best quality of this process lead is claimed to be whiter than the 
"Old Dutch Process" leads and to cover equally as well. 

Krem's or Crem's white is a poorer quality of the "Kremser 
process " lead. 

Kremnitz white is a product from Kremnitz's (German) dry 
precipitation process. 

Flake-white is a pure white lead in a scaly form rather than 
as crystals or grains — the usual form from the Dutch process. It 
lacks opacity or covering power. 

The Clichy or French process is the principal quick-corrosion 
process used in France. The product is known as Ceruse de Clichy. 
It is entirely different from the other decomposition or precipitation 
processes mentioned before. The white lead is formed by passing 
carbonic-acid gas for 12 to 14 hours through a sugar-of-lead or litharge 
and acetic-acid solution, forming a subacetate of lead. The sediment 



QUICK-PROCESS WHITE LEADS, 67 

formed is more or less crystalline, loose or coarse in grain. It takes up 
less oil than the Old Dutch Process leads, allows more light to pass 
through it, hence does not cover nearly so well. 

Greneberg's (German) process consists of the action of carbonic 
acid on finely divided lead and litharge while being rolled constantly 
in tight metallic cylinders. The mechanical friction aids the corro- 
sion at the expense of the purity and durability of the product, though 
there is less exposure of the workman to the corrosion fumes in this 
part of the process. 

Milner's (English) process produces white lead in two days by 
the action of carbonic acid on oxychloride of lead (litharge) by 
grinding them together with common salt in water. 

Pattison's (English) lead is a wet precipitation product — the oxy- 
chloride of lead, made by the action of muriatic acid on galena (lead 
ore). 

The Carter (American) process is a modification. or an improve- 
ment on the Krenmitz (German) process. Metallic lead is melted, 
and while molten is riven into fine particles, like flour, by a jet of high- 
pressure superheated steam. This amorphous powder, of a. steel- 
gray or dark-blue color, is charged into a revolving cylinder 5 to 7 
feet in diameter by 8 to 12 feet long. One end of the cylinder is con- 
nected to an exhaust-fan and the other end to a flue leading from a 
furnace where carbonic-acid gas is generated from burning charcoal. 
Generally the products of combustion from a coke fire under the 
steam-boiler of the planf^-are used for the corroding gas, the furnace 
gases having been washed and purified to free them from any sulphur 
present. The temperature of the revolving cylinder and the charge 
of powdered lead is kept at about 140° F. during the process. Dilute 
acetic acid and hot water are sprayed into the chamber at different times 
during the corrosion process, the stage of which is always accessible for 
inspection by removing samples of the lead without interrupting the 
chemical action of corrosion. The agitation or turning-over of the 
lead and its exposure to the heat is constant during the process. 
About 95 per cent of the lead is changed to white lead by this process 
instead of 60 to 70 per cent by the Old Dutch Process. The presence 
of antimony, bismuth, silver, zinc, and other metals, affects the color 
and quality of the lead by this process as well as in all others. The 
treatment of the white-lead powder after it leaves the cylinder to 
form the dry white lead or the paint paste is similar to the Old Dutch 
or other processes. The powder is repeatedly washed with water to 



OS CARTER'S PROCESS WHITE LEAD. 

free it from the acetic acid, ground in water to a pulp form, and 
floated tlux)ugh a number of tanks and allowed to settle. The greater 
the care used to eliminate the acid the more reliable will be the product. 
The products of all of the above processes, as well as of many 
other quick processes, vary in some degree of quality or form of the 
white-lead atoms from that of the "Old Dutch Process." The latter 
can be said to rank as the standard for purity, fineness, and all other 
qualities which are indifferently imitated in most of the quick-process 
products, the Carter being probably the beet of them. 



Fia. 13. — Hie Carter proceaa for moDufactunng white lead. 

Many of the quick-process leads contain aeetic and carbonic 
acids; the former, being added in excess of the amount necessary 
for a natural rate of corrosion by the old methods, remains in the 
corroded product and requires a more thorough washing to remove 
it than it customarily receives. The acid element is often strong 
enough to redden litmus-paper, which would be discovered if the 
dry white-lead powder could be obtained to make the test. The 
acid causes loss of opacity and rapid chalking. The corrosion of the 
lead is simply removed from the corroding stack to the paint float- 
ing, the carbonic acid and the moisture required in this secondary 
process are both present in the atmosphere, and not only the carbonate 



QUICK-PROCESS WHITE LEADS. 69 

of lead is formed, but this is further reduced to a subcarbonate of 
lead; the latter change constitutes the chalking stage in the decay 
of the coating. Acetic acid and uncorroded lead, left by imperfect 
washing and gidnding, are frequently present in commercial white 
leads. Ten per cent of lead acetate is often found in the "flake- 
whites." Most of the quick-process or impure leads come to the 
market in some form of "whites" with a misleading trade-mark. 

The report of the Committee of Experts appointed by the English 
Home Secretary to investigate the manufacture of lead products, as 
to the character and quality of the product, also their effect upon the 
health of the workmen, was: That they visited forty-six establish- 
ments using various processes for manufacturing white-lead pigments, 
all of which were dangerous to the health of persons so employed, 
and while some of the substitute leads were cheaper to make, and 
possibly a little less injurious to the workmen, their products were 
far from equalling in quality those from the "Old Dutch Process," 
and they could not recommend either the processes or products as 
against the "Old Dutch Process " leads. 

Baryta white is prepared from the native sulphate of barium, or 
from the carbonate of baryta, artificially treated with sulphuric 
acid. (See Baryies, Inert Pigments, Chapter XVIII.) 

Krem's, Nottingham, and Newcastle whites are pure white leads 
differing only in the process by which they are made. Hamburg, 
Holland, and many foreign-made whites contain from 3 to 60 per 
cent of barytes and chalk, and are adiJterated compounds of white 
lead. Venice white generally consists of equal parts of white lead 
and barytes. All pastes and mixed paints classified and marketed 
as "whites" are usually only adulterations of white lead, and no 
responsible and honest corroders of white lead ever so denominate 
their products. The name "white," whatever its trade prefix, 
should usually be viewed with suspicion of its quality. (See tests of 
white lead on the following pages.) 

About 110,000 tons of metallic lead are annually corroded to 
white lead, in the United States, by the various processes, or about 
one-third of the total production of the metal product. 

There are twenty-two manufacturers using the "Old Dutch 
Process" in the United States, and five using the "Quick Process." 

The first white lead corroded in the United States was by Samuel 
Wetherell in 1810 at Philadelphia, followed by Christian Bielen in 



70 ELECTROLYTIC WHITE LEAD. 

1811 at Pittsburg. Cincinnati also had a white-lead plant shortly 
afterward. 

Electrolytic White Lead* 

This process is a radical departure from all of the other processes 
for producing white lead, in not employing acetic acid, but by acting 
upon the lead in the form of pigs with nitric acid, which is generated 
by electricity. The process consists of four consecutive steps: 

First. The electrical preparation of nitric acid and sodium hy- 
droxide. 

Second. The action of the nitric acid on the metallic lead form- 
ing lead nitrate, Pb(N02)2+H2. 

Third. The reaction of lead nitrate and sodium hydrate to form 
lead hydroxide, viz.: Pb(N02)2+2NaOH=Pb(OH)2+2NaN03. 

Fourth. The combination of the lead hydroxide and sodium bicar- 
bonate to form lead carbonate, Pb(OH)2+HNaC03 + NaOH+H20. 

Reactions 2 and 3 may not take place strictly as given, which 
are the theoretical combinations, but some approximate reactions 
are had, for the extra hydrogen present is liberated at the electrode. 

The chemical operations in the process are briefly: 

First, a solution of nitrate of sodium (NaNOj) is decomposed 
by an electric current from a dynamo, the strength of the solution 
not being important — 10° Baimi^ or one pound per gallon suffices. 
This solution is put in a series of wooden cells divided into two com- 
partments by porous partitions. At the positive electrode is fastened 
a pig of lead, at the negative a sheet of copper. On appl)dng the 
current the nitrate of sodium is decomposed according to the equa- 
tion 1 given, nitric acid collecting at the positive electrode and 
sodium hydroxide at the negative. The nitric acid at once attacks 
the lead and forms lead nitrate, which dissolves (equation 2) 
the hydrate of sodium, producing no effect upon the copper at the 
negative pole. 

Finally, both solutions are separately drawn off and mixed, as 
desired, in quantitative proportions in any suitable vessel. The result, 
as shown in equation 3, gives the lead hydroxide as a white 
amorphous precipitate and leaves the nitrate of sodium in solution. 

* "Electrolytic process for the manufacture of white lead." A paper read 
before the American Chemical Society by E. P.Williams. Reprinted in Elec- 
trical World, Sept. 14, 1895, pp. 289-90. Mr. Arthur G. Brown, Inventor, 1892. 



ELECTROLYTIC WHITE LEAD, 71 

This is practically the original nitrate, and can be used over and over 
again as the source for more nitric acid. The loss of the sodium is 
small, and a little additional fresh sodium hydrate restores its strength. 

The lead-hydrate precipitate (Pb(0H)2) is then filtered from the 
sodium hydrate by a rotary separator, and the nitrate of sodium 
returned to the original reservoir. 

The fourth step is in some respects the most interesting of all, 
and consists in adding to the lead hydroxide a solution of bicar- 
bonate of soda (or the normal carbonate). Reaction 4 at once takes 
place. It will be noticed that the sodium hydroxide is the product 
in solution, and lead carbonate the precipitate. 

The sodium hydroxide removes most of the impurities, if there 
are any, in the hydrate of lead. It dissolves any salts of alumina 
or zinc present, and it removes the organic matter. These impurities 
appear in the solution, leaving the precipitated lead remarkably fine 
and white. The hydroxide of sodium is again converted into bi- 
carbonate by passing carbonic acid into it, and this is used again. 
Thus the main agents in each of the two principal steps, the nitrate 
of sodium and the bicarbonate of sodium, are made to do duty over 
and over again with but slight additions to restore the strength. 

The use of free nitric acid in the process is objectionable, as under 
the influence of electricity it breaks up with intolerable fumes; also 
for other reasons. Acetic acid is also objectionable for the same 
reasons, hence the recourse to sodium or potassiimi nitrates for the 
reactions. 

The cost of white lead by this process is but a fraction of that 
by the "Old Dutch Process," as the lead is used as it comes from the 
smelting-fumace in pigs and requires no remelting or casting into 
buckles or shreds, as in the corrosion processes, and the whole process 
is complete in a day, or, for that matter, in an hour, as all of the re- 
actions take place rapidly, if not instantaneously, no free acids are 
used, and the sodium compounds are recovered, as noted. 

The texture of the lead product is almost molecular in fineness 
and does not require grinding, it being so fine that it remains sus- 
pended in the water for a long time, and in order to filter it a special 
brand of cloth is used, as filter-paper would scarcely retain it. 

Its covering power applied side by side with the Dutch Process 
lead appears to be equal to it, possibly a little better, but never found 
to be less. 

Whether the electrolytic lead will displace the "Old Dutch Process" 



72 ELECT BOLY TIC WHITE LEAD, 

lead to any great extent remains for time to determine. The French 
or Clichy process lead, or **Clichy white," was thought at first to be a 
revolutionary one, but the product finally proved to be decidedly 
inferior to the Dutch Process lead, from its crystalline character. 
It does not give the opacity or body, or spread as well under the brush, 
or cover as much surface as the Dutch Process lead. 

The ^' Dutch Process " lead forms a globular atom, viz. : two atoms 
of the carbonate, PbCO,, and one atom of the hydrate of lead, 
Pb(0H)2, but this composition does not always appear to be of con- 
stant quality, as much depends upon the care given during the corrosion 
part of the process. 

Lead hydrate, or the hydrate oxide, is a white amorphous sub- 
stance. The carbonate may be either globular or crystalline, depend- 
ing upon the methods of its preparation. Now, certain qualities 
of these two forms are quite unlike, and this explains why the use 
of one has continued and the other been abandoned as a pigment. 
The atoms of the one form are said to be from ijil-^i^ to ^xi^-^-^ inch in 
diameter, and in the grinding with the oil take it up somewhat as a 
sponge absorbs water. In the " Dutch Process " leads, when properly 
corroded, the atoms are globular, and to this is due the greater body 
and permanency of the paint over that from any of the quick-process 
leads. The crystalline form does not absorb near the same amount 
of oil, no matter how finely it may be ground, as the surface of the 
crystals, either whole, as formed, or crushed in the grinding, are im- 
pervious and do not have the same light-dispersing or reflecting power. 
Hence their poor covering quality; and they do not bond to or in the 
vehicle as well — they save oil. 

If by the electrolytic process it is possible (as it is claimed) to 
produce a pure carbonate of lead, or a mixture of the carbonate 
and hydrate of lead in any proportion required, and the product 
proves to be fine and globular instead of coarse, granular, or crystalline, 
there should be no doubt regarding its merits, but the few hundred 
tons thus far spread do not afford sufficient data for a wholesale 
abandonment of the " Old Dutch Process " with its centuries of 
established reputation. 

Mr. E. Bailey, York, England,* has invented a so-called electrical 
process for the manufacture of white lead and other metallic-oxide 
compounds. An electric arc keeps the lead in a molten state. The 
melted lead is then acted upon by gaseous vapors blown through it 

* London Electrical Engineer^ Jan., 1901. 



ELECTROLYTIC WHITE LEAD. 73 

to produce the carbonate or oxide of lead. The fumes produced 
are blown into chambers having canvas or fine-fabric cloth covers 
or roofs. The fine powders fall down and are collected for pigments, 
and require no washing. The uncondensed vapors escape through 
the cloth screens to the atmosphere or stack. A saving of 50 per cent 
is claimed for the process over that of the " Old Dutch Process," but 
like all other quick-process leads, the merits of the product are yet 
to be established. 

There never has been any difficulty in quickly corroding or oxi- 
dizing metallic lead or zinc for use as a pigment. The difficulty with 
all quick-process leads is in the quality of the product, and, though the 
processes are patented from Dan to Beersheba, the ornate devices and 
claims of the patent-office document do not put the wearing and 
other desirable qualities into the product that the patent claims allege 
to be there. 

There is nothing electrolytic in the Bailey process. The melting 
of the lead by an electric current previous to or during the action of 
the corroding gases produces no electrolysis in the molten metal, the 
heating of which could be as readily and more cheaply done by a coal 
or other fire. The collection of the metallic vapors by screens is an 
old method of pigment manufactiu-e — ^in use for the production of 
sublimated lead, as described in Chapter VI, and is void of electro- 
lytic action. 



CHAPTER VI. 



WHAT CONSTITUTES A GOOD WHITE LEAD. 

Mulder's and Phillips's analyses determined that there are three 
varieties of white lead in the best classes of that product experimented 
with under the general formula of the hydrated carbonate, PbCO,, viz.: 

2Pb''C0,. Pb^'HaO,; 5Pb''C03,3Pb''HA; SPb^'CO,. Hfi^. 

A properly corroded white lead should contain oxide of lead com- 
bined with water — water lead (hydrate of lead, Pb(0H)2.), from 25 to 
32 per cent, which in its effect upon the vehicle is similar to red lead ; it 
absorbs oxygen and hardens the coating by converting a part of the 
oil into a soap that has no covering power whatever. The other 75 
to 68 per cent should be oxide of lead combined with carbonic acid 
(carbonic-acid lead), that really injures the oil in the paint, but gives 
all of the covering properties that the paint possesses. The chalking 
of white-lead paint is due to this 75 per cent of carbonic-acid lead. 
A paint composed wholly of carbonic-acid lead will, in a short time, 
chalk as freely as a whitewash. The carbonic-acid lead gives the 
whiteness or color, the water lead, the hardness or durability to the 
coating. 

In the following table from Heiss's experiments, 

Number 1 is a good dense white lead. Specific gravity, 6.32 

2 " dr>' white lead. " " 6.50 

3 " crj'stalline transparent lead. " " 6.05 



(t 



tt 



Oxide of lead. . . 
Carbonic acid. . , 
Combined water, 



Number 1. 



85.95 per cent. 
11.14 " " 
2.91 " " 



Number 2. 



86.18 per cent. 
10.44 " " 
3.38 " " 



Number 3. 



83.53 per cent. 
15.70 " " 
0.77 



tt 



it 



Another comparative table is: 




White lead, best quality. . 

"2d " .. 

"3d " .. 

Residue lead 

Improperly corroded or 
useless lead 



86.80 percent. 
86.24 " " 
86.03 " " 
84.69 " " 



83.47 



tt 



tt 



Carbonic Acid. 



11.16 per cent. 
11.68 " " 
12.28 " " 
14.10 " " 



16.16 



tt 



tt 



Combined Water. 



2.00 percent. 
1.81 " 
1.68 " 
0.93 " 



tt 
tt 
tt 



0.25 



tt 



tt 



74 



WHAT CONSTITUTES A GOOD WHITE LEAD. 



75 



In Prof. Hurst's analyses of four samples of "Old Dutch Process'' 
white lead, the carbonate ranged from 65.35 to 72.15 per cent, the 
hydrate from 25.19 to 36.14 per cent, and the moisture 0.42 per cent 
to nil. 

Mr. Converse's analyses of five samples of the best brands of 
American "Old Dutch Process'! leads gave 



Lead carbonate. 

" hydrate. . 

" oxide. . . . 
Water 



85.32 
14.83 

• • a • • 

0.03 



79.37 
19.80 

'6!2i 



78.68 

20.11 

1.48 



77.98 

20.60 

1.48 



69.96 
30.19 

'6!67 



A very hard white lead that contains no water lead will not harden 
when spread, but brush off like a lime whitewash. When this occurs 
where the necessary amount of oil has been used, the painter can be 
quite sure that the lead is a poorly corroded or quick-process lead, or 
a synthetically formed lead from an acid-solution process. Carbon- 
ate of lime (chalk, whiting), gypsiun, and other so-called inert sub- 
stances added to such a lead do not correct this chalking; they only 
disguise it for the time being, and increase the tendency of such lead 
paints to turn yellow and lessen their covering power. 

All white lead on external exposures is liable to chalk, because it 
contains too much carbonic-acid lead as it comes from the manufac- 
turer, or has taken another portion of carbonic acid from the atmos- 
phere at the expense of the water lead, and formed a subcarbonate of 
lead (the chalk product, PbCOj), or has too little oil in it when spread, 
or has not been well ground in the oil. Paints thus affected, if brushed 
over with a coating of white lead ground with an excess of oil, will prove 
to be more durable and less affected by future chalking than the 
original coating, or a new heavy coat from the same lead. 

The formation of a lead soap in the ordinary process of grinding 
and mixing white-lead pastes or paints is a disputed point by paint 
chemists. But the lead hydroxide and the free linoleic acid in linseed- 
oil, if acetic acid is present in the white lead, vnll combine and form a 
lead-soap mixture. The paint containing this soap, on exposure to 
the weather, soon loses its lustre and will crumble or chalk. 

The presence of lead soap in many, if not the most, of white-lead 
paints is shown by Prof. Church, in his "Chemistry of Paints and 
Painting," as follows: "Upon a piece of glass place a small quantity 
of the white-lead paint, and add a 10 per cent solution of sulphuric 
acid. Work the mixture together with a glass rod or spatula into a 



76 CHAIKING OF WHITE LEAD. 

cream-like condition. The acid will soon destroy the oily or hydro- 
fuse character of the paint. With zinc white or baryta white in oil 
no such admixture is possible, for in these paints the oil will not 
saponify owing to the absence of an acid." 

Mulder's and other experiments proved that the chalking of white 
lead was due solely to the absence of water lead in the pigment, unless 
the lead was badly adulterated, in which case this effect was directly 
traceable to the adulterant used. 

In general, all the latter-day-process white leads are more inclined 
to chalk than the "Dutch Process" leads. This arises (as stated 
before) from the smaller amount of water lead in their composition, 
and bemg of a crystalline instead of a globular form. In the grinding 
process these crystals or grains are broken down, and the combined 
or formative water necessary for their existence in the form of crystals 
is dispersed, rendering the broken lead atoms more sensitive to the 
attack of the atmospheric carbonic-acid element, that finds in their 
sharp angular form a more favorable surface to act upon. Only a 
comparatively slight action of the carbonic acid on the freshly crushed 
atoms of the lead is required to change it to the subcarbonate, and 
leave it free to be brushed off by friction or washed out by storms. 

Old white lead, or that which has been ground in oil to a paste for 
a year or more, chalks decidedly less than recently corroded lead 
by whatever process it is made. The atoms of the pigment in the case 
of old lead have had time to release themselves from the tension due 
to their formation and grinding. Adulterations do not prevent these 
inexorable chemical changes in the lead pigment; they only increase 
the disintegrating action. 

If oxide of zinc is used to give hardness in place of the water lead 
m pamts exposed in open air, the atmospheric moisture and carbonic 
acid changes the oxide of zinc (ZnO) to a zinc carbonate (ZnCO ) 
whose volume is nearly double that of the oxide from which it was 
formed in the hardened paint, and peeling takes place instead of 
chalking The cheaper forms of zinc oxide contain zinc sulphite 
which blackens the paint and otherwise injures the coating 

If gypsum, barytes, or silica are used for the adulteration, thev 
lack both m covering and light-dispersing (coloring) power, and. from 
their sharp, angular, or irregular form as pigment-atoms, do not bond 
themselves m the oil, for which they have no affinitv, nor enter into 
combmation with, or bond to the covered surface"! They prevent 
chalking only as their presence leaves less white lead to chalk even 



CHALKING OF WHITE LEAD. 77 

if they do not actually increase its tendency towards that change, as 
the sharp, rough surface of the paint containing these adulterants 
holds the atmospheric tnobture and gases closer and longer for their 
action. {See Decay of Paint and Inert Pigments, Chapter XXVII.) 

White lead unites thoroughly with the oil. Zinc white combines, 
but very slowly; barytes does not combine at all. All pigments that 
contain crystals or are granular are deficient in the iight-dispersing 
power, even if they have the spreading quality. The granular char- 
ficler of quick-process white lead is its great weakness. 



Fio. 14. — Effects of Bulphuroua gases on white lead, 
Backiroun.l in linf white. BftckBroumf in wliite Ibbij. 

The quality of the vehicle has much to do with chalking and the 
decay of white lead. A good linseed-oil will better preserve a poor, 
or an adulterated white lead, than a poor oil could the best of white 
lead. 

Pure white lead is soluble in dilute nitric acid. A sample treated 
with thb reagent should pass entirely into solution, leaving no residue. 
If the sample is in the form of a paste or paint the oil can be removed 
by washing it with benzine or ether, and the powder, when dried upon 
blotting-paper, should leave no stain. 

The action of sulphuretted hydrogen upon white lead is that the 
sulphur element unites with the lead to form the sulphide of lead (PbS), 
â– which is a dark color. Further exposure changes the sulphide to a 



78 ADULTERATION OF WHITE LEAD. 

sulphate (PbSOj), which is white and of good covering power. Hence 
the full change would not be detrimental to the coating in point of 
color were not these changes in the dried coating attended by a change 
in the volume of the lead-atoms at each step in their progress from a 
carbonate to a sulphate that ensures a forced disintegration of the 
coatings. (See Decay of Paint, Chapter XXVII.) 

The cause of white-lead paint, when spread over darker colors, 
deteriorating after a short period and showing dark, is not alone 
owing to the changes mentioned above, but also that the fatty acids 
in the oil gradually expel the carbonic acid in the lead-atoms and form 
a clear lead soap, through which the darker colors beneath show. 

Adulteration of White Lead, 

The adulteration of white lead under the forms of a paste or 
mixed paint has reached a point where the multitudinous trade- 
marks under which they are marketed are positively of no value 
to determine their quality; the only safety lies in purchasing from 
responsible business firms of national reputation for the standard 
quality of their products and business methods. 

A late examination of commercial white-lead pastes purchased 
in the open market resulted as follows: 

Seventy-five different white-lead pastes, under 29 different trade- 
marks and symbols, embellished with 14 qualifying adjectives, and 
made by 17 different manufacturers, plus 1 unknown firm that 
furnished 14 different brands, were analyzed by 16 different anal)rsts. 
The condensed results are, viz.: 

Sixteen had no white lead in their composition, but were mixtures 
of barytes, silica, gypsum, zinc oxide, or whiting in some proportion 
of three or more of these substances. 

Fifty-nine of the samples had white lead from 1.24 per cent to 
47.62 per cent, and averaged for all 23.35 per cent of lead. 

Seventy of the samples had barytes from 15.60 per cent to 86.37 
per cent, and averaged for all 49.61 per cent. One or more of the 
above adulterants was used with white lead of uncertain quality for 
the balance of the paste. 

In 5 of the samples silica replaced the bar3rtes, otherwise the 
usual group was unbroken. 

Seventy-five of the samples had oxide of zinc from 59.20 to 
3.60 per cent, averaging for all 27.48 per cent. A free use of gypsimi 



ADULTERATION OF WHITE LEAD. 79 

and whiting aided the white lead to give the paste a semblance of a 
pure white-lead product. 

Thirty-two of the samples had sulphate of zinc in the paste from 
84.85 to 0.88 per cent, averaging for all 22.09 per cent. 

Five samples had chalk from 24.30 to 0.85 per cent, averaging for 
all 9.60 per cent. 

Wherever carbonate or sulphate of lime was used it was at the 
expense of the white lead. 

The adulteration of all of the brands, other than the 16 that had 
no white lead, ranged from 99 to 44 per cent, and averaged for the 
69 pastes 80.4 per cent. 

In 5 of the samples oxide of zinc was mentioned as constituting 
a small part of the paste, but nothing was said about the 30 to 40 
per cent of barytes in their composition, or of a like amount of the 
other adulterants, or the snudl amount of white lead. 

One sample had a notice that $1000 would be paid if the white 
lead in it was not pure. It had no white lead in it. 

One sample offered $100, same conditions as above. It was of 
the same character. No white lead in it. 

One sample had a $250 penalty for any imitation of the lead in it, 
or of the label on the package. • It had 99 per cent of adulterants. 
Possibly a piece of the label might have fallen into the paste before 
analysis of it. 

Probably the above adulteration record represents the larger num- 
ber of commercial white-lead and tinted-color paints in the market. 
Unfortunately, most of them get juggled off, and generally at the 
price of really first-class paints; but so many people are willing to 
be humbugged to save a few cents on a pot of paint that it may be 
a mistake to enlighten them on the subject. What the character and 
quality of the oils are that form an essential part of all these adul- 
terated and misleadingly named compounds, one can readily imagine. 
Surely they are equally deceptive and uneliable. 

Maize-oil is used frequently to the amount of 25 per cent in the 
grinding of paint pastes and paints to prevent them from settling 
in the package. It is a non-siccative oil, and while not exerting any 
materially bad influence in a paint, its use with linseed-oil in any 
quantity can only be considered as an adulterant. Its cost is about 
one-half that of commercial linseed-oil. Its use requires more driers. 

The use of 10 to 15 per cent of barytes in a white-lead paint may 
not be particularly objectionable in point of durability, if the lessened 



80 WHITE LEAD AND ZINC PAINTS, 

cost is the object desired. Lead pigments do not cover so well with 
barytes, but zinc oxide covers better; barytes gives weight that the 
zinc is deficient in. Some master painters advocate the use of barytes 
for the reason that it brightens dark colors and saves oil; but ignore 
the fact of its darkening light colors, also its tendency to yellow them 
from the sulphur element in its composition and its general deficiency 
in covering power. The advantages in any case do not warrant the 
use of from 50 to 80 per cent of this or any other substance which does 
not unite with the oil, and of themselves are unfit for a paint. As a 
rule, no responsible paint firms will offer such paints under their own 
names. 

Zinc oxide is often mixed with white lead for other than adultera- 
tion purposes. It is added to correct the tendency of pure white 
lead to turn yellow from the action of sulphurous fumes in the at- 
mosphere. It makes the paint harder, and possibly prevents "chalk- 
ing" to some extent, as there is not so much lead to chalk, but adds a 
tendency to peel on outside exposures — a, result worse than chalking. 
It is difficult to get a homogeneous blending of the two pigments so 
individually imlike in character, owing to their different processes 
of formation. Their mechanical association is far unlike that of a 
neutral whole, even with the best of supervising care in blending them. 

Many of the difficulties in mixtures of white lead and zinc oxide 
are overcome by the use of a natural admixture of the lead and zinc 
in the form of a pigment known as sublimed lead or white paint, 
hereafter described. 

White Lead vs. Zinc Paints. 

Mr. G. R. Henderson, of the Norfolk and Western Railroad, 
reports a series of exposure tests to determine the efficiency of lead 
and zinc paints. For the different materials he reaches the following 
resiQts: 

*' Tin. The best results were obtained with the first coat of white 
lead and second coat of white zinc. The second coating of zinc gave 
generally the best results, and the second coating of lead the worst. 

" Galvanized Iron. The same remarks apply to galvanized iron 
that are given for tin. 

'^ Sheet Iron. The mixture of one-third white zinc and two-thirds 
white lead, for both coats, gave the best results on this material, and 
in general the zinc paints gave better results than the lead paints. 

*' Poplar. The second coats of zinc showed up well on poplar, no 



TESTS FOR WHITE LEAD. 81 

matter whether the priming coats were white lead or white zinc or 
mixed lead and zinc. The lead second coating showed up the worst on 
this material, but in each case where the second coat was of zinc, 
totally or partially, the paint was in a perfect condition. 

" White Pine. The remarks made relating to poplar apply to white 
pine also. 

" Yellow Pine. This material seems to be difficult to properly treat 
with paints; the best results were obtained with the first coat of lead, 
and the second coat of lead and zinc mixed. Where the first coat was 
of lead and zinc mixed, or entirely of zinc, the results were poor 
throughout, which seems to indicate that as a general thing the lead 
is better for priming on this material. 

" Conclusion. The lead priming and zinc coating is generally good 
for tin, galvanized iron, poplar, and white pine. Sheet iron showed 
up best with both coats of mixed paints. Yellow pine appeared best 
with the first coat of lead and the second coat of lead and zinc mixed. 
Comparing the materials which were painted, we find that generally 
poplar retains the paint better than white pine, and would, therefore, 
be preferred for siding on buildings, etc. Yellow pine seems to be 
the worst of all for this piurpose. Black iron, as a whole, seems to 
retain the paint better than either tin or galvanized iron." 

Tests for White Lead. 

There are many tests for the purity of white lead, a simple one 
being to put a globule of the paste or dry powder in a cavity formed 
in a piece of charcoal, and expose it to the heat of a common blow- 
pipe, readily extemporized in any shop. If even 10 per cent of adul- 
terants are present it will not be possible to melt the mass. 

White lead is tested for barytes by dilute nitric acid, in which 
barj^tes is insoluble, while the white lead passes completely into the 
solution. Whiting and chalk are detected by the nitrate solution 
yielding a white precipitate with oxalic or sulphuric acid or oxalate 
of ammonia, after having been treated with sulphuretted hydrogen or 
a hydrosulphuret to throw down the lead. 

White lead dissolves with efferv^escence in hot hydrochloric acid as 
chloride of lead, which crystallizes into needles on cooling. Dilute 
nitric acid easily dissolves white lead with effervescence and an escape 
of carbonic-acid gas. White lead heated on a knife or piece of metal 
turns yellow. Sulphuretted-hydrogen fumes in the air turn white lead 
gray. 



82 



TESTS FOR WHITE LEAD. 



Substance. 


Conduot towards 


Heated before the 




Muriatic Acid. 


Caustic Soda. 


Blowpipe. 


Whiting or chalk 
in any form. 


Soluble with 
effervescence. 


Unchanged. 


Becomes incan- 
descent and turns 
tumeric brown 
after cooling. 


Commercial white 
leads. Pearl and 
other whites. Car- 
bonate of lead, etc., 
quick-process leads. 


Soluble with 
effervescence and 
deposition of 
small crystals. 


Soluble with- 
out residue, or if 
of poor qualit;^. 
80 per cent will 
only be dissolved. 


Coating formed 
on the charcoal 
is citron-yellow 
while hot. sulphur^ 
yellow wnen cold; 
is easily fused with 
metalhc beads. 


Pattison's white 
lead. Lead oxy- 
chloride. 


Same as above. 


Same as above. 


Same as above. 


Zinc white or 
oxide. 

• 


Soluble, no ef- 
fervescence. 


Soluble with- 
out residue. 


Yellow while 
hot, white when 
cold. 


Antimony white. 
Antimonious- acid 
pigments. 


Same as above. 


Same as above. 


White, easily 
volatilized, me- 
tallic globules 
which give oflf 
white smoke. 


Bone^ash, Bone- 
black. PhoG^hateof 
calcium, Ca.(P04)a. 
(Basic steel furnace 
dag.) 


Soluble after 
heating. Effer- 
vescent at first. 


Unchanged. 


Unchanged, but 
becomes incandes- 
cent. 


Barytes, Blanc- 
Fixe, mineral white, 
and other pigments 
of the native sul- 
phate of barium. 


Unchanged. 


Unchanged. 


After ignition, 
if moistened witli 
muriatic acid, 
gives off sulphur- 
etted hydrogen 
vapor. 


Gypsum (Hy- 
drated sulphate of 
calcium). 


Unchanged. 


Unchanged. 


Incandescent, 
like barytes. It 
heated in a tube, 
gives water vapors, 
etc. 


China, pipe, pot- 
ters', ana oiner 
days. 


Unchanged. 


Unchanged. 


Incandescent, 
same as barytes 
and gypsum. If 
moistened with a 
solution of cobalt, 
and heated by the 
blowpipe, turns 
blue. 



SUBLIMED LEAD. 83 

It is not generally known that dry white lead and oil combine with 
such energy that if linseed-oil is poured upon a large quantity of the 
lead, and the mass is allowed to stand a few hours, the temperature of 
the mass becomes so high that the oil is carbonized and colors the mass 
dark or even black. All white-lead paste is liable to turn brown if 
exposed freely to the air, hence should be kept closely covered, or 
water be kept on the paste when the package is opened. 

Gmelin states "That the ' Old Dutch Process ' white lead diffused 
through water, under the microscope appears as non-crystalline, 
round, and oval globules, .0001 and rarely .0003 or .0004 of an inch in 
diameter, while in the quick-process leads the globules are distinc- 
tively larger and more transparent and crystalline." 

Dry white lead is tested by heating 100 grains red hot and stirring it 
for a moment. Its loss in weight by driving off the carbonic acid should 
be from 13 to 16 grains. If more or less loss is incurred, the lead is 
probably adulterated, and should be submitted to further test to deter- 
mine the character of the adulterant, as shown in the table opposite. 

Svblimed Lead, PbSO^. Specific Gravity, 6.258. 

Sublimed lead is a by-product obtained in the smelting of non- 
argentiferous lead ores. It is known in the trade as Joplin lead, from 
its place of manufacture, Joplin, Mo.; also as Picher lead, from the 
name of the manufacturing company. It is made in two colors — 
white, suitable for all purposes that the hydrated carbonate of lead 
18 used for; and blue, which is a preferable color when used as a paint 
for iron. Both colors are used in the manufacture of india-rubber 
articles. The chemical composition of sublimed lead is sulphate and 
anhydrous oxide of lead, both amorphous; the good qualities of the 
white are also present in the blue-colored product. There is a small 
percentage of zinc in the Missouri lead ores, which in the process of 
smelting is converted into zinc oxide and is found in the sublimed-lead 
product, as may be seen in the analyses of a sample of the white 
product given on page 84. 

The blue pigment owes its color to the lead sulphide and car- 
bonaceous matter from the bituminous coal used as a fuel in the 
smelting-fumace. 

Sublimed lead is prepared in special furnaces, the process being 
patented and known as the "Lewis and Bartlett Bag Process for 
collecting lead fumes." The process has been mentioned and described 
in more or less detail in the following papers. 



84 



SUBLIMED LEAD. 



Analyses Substances. 

Sublimed lead, PbSO^ 

Protoxide of lead, PbU . . . . 
" " zinc,ZnO 

« " iron, Fe,0,- • • • 

Calcium oxide, CaO 

Carbonic acid. CO, 

Sulphuric acid, SO^ 

Insoluble matter 

Water 

Loss 





Averages. 




Substances, Per Cent. 


65.46 
25.86 


to 


65.00] 
25.89 


P^f ^/?'- [ Metallic lead, 71 .831 


5.96 




6.02 




' Metallic zinc, 4.808 


0.03 
0.02 




0.03 
0.02 




; Oxygen com- \ ^ -^^ 
' bined, -W.50» 


1.53 




2.00 




' Carbonic acid, 1.765 


0.04 




0.00 




* Iron, lime, and in- 

* soluble matter, 0.130 


0.08 




0.08 




0.86 




0.69 




* Sulphur, water. 


0.19 




0.27 




' and loss, 0.958 




100.00 



Mineral Resources of the United States, 1883-4, p. 427; Engineer- 
ing, 1884, p. 495; Engineering and Mining Journal, Vol. XL, 1885, 
p. 4; B. und H. ZeUung, Vol. XL VII, p. 346 (describes the works of 
the Bristol Sublimed Lead Company, Bristol, England, where the 
process is in operation) ; Prera^ss Zeitsch, Vol. XVIII, p. 195; Fresenius 
Zeitsch, Vol. VIII, p. 148; Transactions of the American Institute of 
Mining Engineers (Washington meeting, February, 1890), illustrated. 

The galena ore is first smelted with coal and slaked lime in the 
special furnace, using an air-blast to obtain the required heat, about 
800® F. ; the hotter the fire the more lead is volatilized, and the more 
"fume" is produced. 

The philosophy of the process is that galena ore, or the native lead 
sulphide, when heated to nearly a white heat, vaporizes slowly, and 
the vapors in contact with an* excess of air passing through the fur- 
nace bum into lead sulphate. Simply heating a mass of galena ore 
does not, however, form the sublimed-lead product. When the ore 
is properly roasted in the special furnace the temperature required 
to effect the change to sublimed lead is far below that required in the 
vaporization of lead sulphide. The ore when rapidly heated, even if 
it does not actually vaporize, has a tendency to do so, which is suf- 
ficient for the sublimation process, and under favorable conditions will 
bum at a dull-red heat. 

The process is the same in principle as that used in the collection 
of the oxide of zinc. The use of "fume" from the smelting of lead 
for a pigment is very old. Bishop Watkins, in his scientific works 
in 1778, mentions the use of gray fume for a paint; but its color was 
then objectionable as against the white product of corroded white 
lead by the "Old Dutch Process." 

The direct fume-product from the combustion of galena ore is not 



SUBLIMED LEAD, 35 

yet sublimed white lead. The products of the smelting-furnace are 
pig lead, pasty slags containing more or less lead, zinc, and other 
metallic constituents of the galena ore and "fume." The latter is 
drawn oflf by an exhaust-fan through a settling-chamber to a 
bag-house which contains a large number of woollen bags for 
filtering the fume out of the combustion gases. This " fume " is a 
lead-colored, impalpable powder known as "blue powder." It is 
ignited and allowed to burn for several hours, which converts it into 
white coherent crusts. These crusts, with some oxidized ores, set- 
tling in the flues and hearth slags, are next charged into a special 
furnace and exposed to a very hot coke fire. The products of this 
smelting are pig lead, slags poor in lead as waste material, and the 
"fume," which is now a perfectly white impalpable pigment suspended 
in the air. This is drawn through a series of cooling flues, where a 
further purification takes place, a part of the product settling and 
carrying down small quantities of any impurities that have escaped 
the action of the heat. The sublimed lead is now arrested by forcing 
the gases and lead into strainers of fine textile fabric, where the gases 
escape by filtration. The sublimed lead is taken from the strainers 
and is ready for the market. 

The process assures a more intimate combination of the vaporized 
atoms of the lead and zinc to form a neutral whole in which every 
atom is approximately of the same physical character and equally 
affected by any of the causes that injure a paint, than is possible 
in a pigment formed by a chain of partly chemical and partly mechani- 
cal operations acting upon a number of separate substances at different 
stages of the manufacturing process. Sublimed lead is absolutely 
free from soluble acids or sulphur, is amorphous, non-crystalline, and 
fine and smooth in the character of its atoms. Being a pyrogenic- 
formed substance, it is not affected by heat or deleterious gases of 
the atmosphere or manufactories ; does not turn gray on long expo- 
sure to the sunlight, and is not liable to chalk like corroded white lead. 
Though it dries firm and of almost ivory hardness it does not blister, 
crack, and peel like the oxide of zinc or mixtures of zinc oxide and 
corroded white lead. It is elastic, mixes thoroughly with the oil, and 
dries well without an excessive use of driers, either metallic or liquid. 
Pound for pound it covers better than white lead, and keeps its color 
better, as the following comparison of painted boards exposed to 
furnace gases shows. The action of the weather upon sublimed lead 
18 confined to the surface destruction of the oil vehicle. Rain and 



86 SUBLIMED LEAD. 

wind do not affect or remove the pigment. When the surface of the 
paint shows decay, the coating can be repainted without removing 
the old paint. Either the white or blue products of sublimed lead 
will take a tint more uniformly than is possible with any mixtures 
of white lead and zinc oxide incorporated by the usual grinding proc- 



FiQ. IS. — Sublimed lead. Fio. 16, — Pure corroded white lead. 

The cuts are reproduetiona from a photograph of a picket fence. 
each alternate picket being painted with the different leads at the 
same time and by the same painter, using the same oil as a thinner 
for the pastes, and the same driers, a separate brush being used with 
each paint. Three coats of each paint were applied, each being 
allowed to dry thoroughly hard before the next coating was spread. 
After an exposure of three years and one month the photoftraphs 
were taken. The cuts are magnified ten times. The chemical reac- 
tion between the corroded white lead and the oil forming a lead or 
paint soap is clearly apparent. The shrinkage of the paint soap 
has caused the coating to crack. Moisture has entered and loosened 
the bond of the paint to the covered surface, white the soluble char- 
acter of the lead soap has caused the whole coating to "craze," and 
it is ready to fall off by any slight mechanical disturbance. 

The sample of sublimed lead shows that no reaction (or but 
little, if any) has taken place between the sublimed lead and the 



SUBLIMED LEAD. 



Fia. 17.— Sublimed lead. 



FlQ- 18. — Pure corroded white lead 



88 SUBLIMED LEAD. 

otH, and that the impervious character of the coating has kept out 
the moisture, and the paint is still firmly bonded to the wood; the 
streaking being due to the running down of the road dust when wetted 
by storms, no crazing being noticeable. 

Fig. 17, sublimed lead, and Fig. 18, pure corroded white lead, are 
two-coat applications of the respective paints on a well-seasoned 
board after an exposure to the atmosphere for three years. 



CHAPTER VII. 

ZINC OXIDE (ZnO) AND OTHER ZINC PAINTS. 

Metallic zinc (Zn), specific gravity, 6.86 cast; 7.14 to 7.20 rolled - 437.5 
pounds per cubic foot; combining weight, 65.4; tensile strength, 5000 to 6000 
pounds per square inch; electrical conductivity, 29. It melts at 780^ F. and 
begins to volatilize in the open air at 800"" to 825"* F. 

Zixc is electro-positive to copper and iron, whether in solution or 
in contact. In contact with iron or steel it forms a galvanic pile and 
decomposes with the evolution of hydrogen. It is used in this form 
to protect steam-boilers from corrosion. Each pound of zinc decom- 
posed evolves 5.6948 cubic inches of hydrogen that weigh 210.29 
grains and develop an electrical or decomposing energy of 1.172 
horse-power when used in a sulphuric-acid battery or 1.06 horse- 
power in a Bunsen or Grove battery. 

This electrical energy, while not so strong as in the oxide or salts of 
zinc in a battery form, is still present, and is recognized by painters 
as "movement in the paint" which is very marked with zinc pig- 
ments, as will be hereafter explained. 

Mallett's experiments determined that copper and zinc plates in 
contact with iron increased the corrosion of the iron 60 per cent. 
Copper alone in contact with iron, 40 per cent. Eisner found that 
the oxides of tin, zinc, and iron used together in a paint set up gal- 
vanic action enough to crystallize the tin into flakes, which could then 
be rubbed off. 

Zinc is associated with nearly all of the other metals as a mineral 
ore. It is roasted in special furnaces and by processes similar to 
those used in the reduction of lead ores for the metal or its oxides. 
The presence of these associated metals affects the quality and color 
of the oxides. 

The native carbonate of zinc ore, (Zn.CO,) (calamite, zinc spar), 
specific gravity 4.45 to 5.0, is found in many parts of the world in 
heavy beds as crystals, translucent when pure, but tinged more or 

89 



90 ZINC OXIDE. 

less gray, green, or brown, according to the other mineral substances 
associated with it. 

Zinc fonns but one oxide, ZnO, composed of zinc 80.344 per cent 
and oxygen 19.656 per cent; specific gravity ^ 5.42. 

Zinc oxide is produced by two methods: First, by the oxidation 
of the metal (French process), which gives a more dense and harder 
pigment than that prepared by the second method — the sublimation 
of zinc ore (American process). The product from the first process is 
preferable for straight oxide-of-zinc paint; the second is better for 
use in combination with other pigments. Both varieties grind and 
mix better with poppy-seed oil than with any other vehicle, and man- 
ganese-borate drier should be used in preference to lead driers, which 
are liable to blacken the paint on exposure. 

In both processes the zinc is reduced to a vapor b}'' the furnace 
heat (about 850° F.), and is exposed to a current of air that changes it 
to a flake, filament, or needle form, according to the care exercised in 
the process. Some of the particles of the metallic zinc are apt to be 
carried over in the vapor imchanged, mixed with the atoms of carbon 
from the reduction fire. These carbon-atoms tend to give a gray 
color to the product, while the metallic-zinc particles are subject to 
an attack from the carbonic acid in the atmosphere, forming the zinc 
carbonate (ZnCO,) ; this latter change occurring in the pigment after 
it is mixed into a paste or paint. 

The French brand of zinc oxide (La Vielle Montague Co.) is the 
best in the world, probably, due to the purity of the metallic zinc, 
the care exercised in its reduction, and by the use of poppy-seed oil, 
with which the oxide is ground immediately after its formation. 

Most of the American zinc ore contains lead, tin, antimony, bis- 
muth, silver, etc., the oxidation of which in the reduction furnace 
produces white pigments, but they are all blackened by sulphurous 
hydrogen, which aflFects the quality of the zinc product, and aids in 
setting up the electrolytic action that all zinc substances are sensi- 
tive to, as before stated. Sulphur is also present in many zinc ores, 
and causes a yellow color in the oxide. 

History of Zinc Oxide, 

In 1781 a French chemist discovered the process of reducing zinc 
to an oxide, and advised its use instead of white lead, but no special 
results followed. 



HISTORY OF ZINC OXIDE. 91 

In 1796 an Englishman patented a zinc-combination pigment, but 
it did not come into general use against white lead. 

In 1844 Leclair, a Frenchman, made zinc oxide and founded the 
La Vielle Montague Zinc Ck). Leclair died in 1872. He received a 
gold medal from the Society for the Encouragement of the Arts, and 
was decorated with the Grand Cross of the Legion of Honor for having 
improved the practice of painters. 

Leclair disguised the fact of the use of zinc oxide as a pigment. 
His product was always sold as white lead, a precedent that mixed- 
paint manufacturers of the present day follow but too well if the 
analyses of commercial white leads given in Chapter VI are any indi- 
cation. 

The Leclair process consists of volatilizing the metallic zinc in a 
retort, the vapors as they issue from it being met and mingled with a 
current of hot air from a blower which completely oxidizes them. 
The resulting products of combustion are led through a series of flues 
and chambers, where the zinc oxide is deposited in the form of a 
flocculent, impalpable white powder ready for use as a pigment. 

The American process for the sublimation of zincite ore into 
zinc oxide was invented by Mr. Samuel T. Jones in 1850, who con- 
structed a furnace for effecting this purpose. This process was 
improved by Col. Samuel Wetherell for the purpose of working the 
franklinite ore from the New Jersey deposits. 

Wetherell's invention embraced a special furnace and a process. 
The zinc ores are mixed with pulverized anthracite coal, and charged 
into a closed furnace having a perforated grate, through which an 
air-blast furnishes the air necessary for the combustion of the coal 
and oxidation of the zinc. The vapors from the furnace are led 
through a number of flues and chambers, where the coarser particles 
are deposited, while the fine air-floated atoms of the zinc oxide pass 
on and are collected in a number of fine muslin bags, through which 
the combustion gases filter away to the atmosphere; the latter part 
of the process being similar to the Lewis and Bartlett bag process, 
used in the production of sublimed lead, described in Chapter VI. 
The franklinite and zincite ore found near Mt. Sterling, N. J., now 
furnish most of the American zinc oxide. 

About 44,000 tons were produced in the United States in 1900, 
and its use for house-painting is increasing at the rate of about 8 per 
cent yearly. The New Jersey deposit of zincite is almost absolutely 
free from lead, antimony, sulphur, and other metals that affect the 



92 QUALITIES OF ZINC OXIDE, 

color and quality of zinc oxide. In France its use has almost super- 
seded white lead for interior house-painting, the Government pro- 
hibiting the use of white lead for this purpose. 

Zinc oxide made from the ore is used more extensively than that 
made from the metal. The latter not only dries harder and is 
more brittle, but on large surfaces the difference in the whiteness 
of the coatings is very apparent and in favor of the mineral oxide. 
Zinc oxide mixed with water tends to collect in lumps or masses. 
It should be thoroughly dry before being ground with the oil. It 
does not unite so thoroughly with oil as lead or iron pigments, 
nor dry as quickly. 

One-third of one per cent of litharge, added to the linseed-oil 
in which zinc oxide is ground, renders the paint more elastic and 
less liable to peel. 

Mixtures of zinc oxide, white lead, and ground silex-barytes 
in the proportions of about one-third each, prove very durable in 
southern climates and seacoast exposures. The silex gives body 
to the paint, but being transparent, detracts from its coloring or 
covering power. There is less white lead in the mixture to saponify 
with the linseed-oil to form a lead soap that is quickly washed away 
by storms. The oil also dries out and the white lead is rendered 
liable to chalk. Such coatings cost less than pure white lead, are 
more bulky, and by a little extra labor on the painter's part can be 
made to cover more surface than white lead and zinc oxide, and 
they save oil. 

Zinc oxide 50 per cent, white lead 25 per cent, and Blanc-Fixe 
25 per cent, also stand southern and seacoast exposures better than 
white lead alone. When the above mixtures are used on wooden 
structures the barytes and silica act as fillers in the first coat, and the 
percentage of either substance can be greater than in the other coats. 

Pure zinc oxide is a pure white pigment but little affected by 
sulphur fumes, and does not yellow the oil with which it is ground 
or mixed. It is of itself a good drier, and is used in the preparation 
of kettle-boiled oil. When it forms the principal pigment in a paint, 
other driers are detrimental, as the coating has a tendency to harden 
upon the surface only, and remain viscid below and peel readily. 
Linseed-oil dries harder than poppy-seed or walnut-oil. In zinc- 
oxide and white-lead mixtures, more lead than oxide will be required 
when linseeil-oil is the vehicle, in order to keep the coating elastic 
and avoid its tendency to "craze" or peel. 



QUALITIES OF ZINC OXIDE. 93 

Zinc oxide carries more oil than white lead, hence spreads better 
and reduces the tendency of the lead coating to chalk, simply be- 
cause there is less lead in it and the atoms of each are better coated 
with the oil from the larger quantity of it necessary with the zinc 
pigment. Mixed-zinc oxide and white-lead coatings are, in general, 
more durable than either coating alone, provided both the pigments 
are pure. As zinc oxide costs more than white lead, the adulteration 
of it is quite as general, and with the same substances. (See Chap- 
ter VI.) 

It deteriorates by long keeping and loses much of its covering 
power, which can be restored by heating the pigment. When freshly 
made or heated and exposed to moist air, it changes, by absorbing 
carbonic acid, into the carbonate of zinc (ZnCO,). Painters say 
''it spoils," and guard against this change by closely covering or 
sealing it, or mixing it at once into a paste or paint, where the oil 
protects it, except those particles that lie upon the surface. The 
carbonate of zinc so formed is crystalline, loses density, and hardens 
so that it cannot be pulverized or ground without extreme effort. 

In this change 81 parts of zinc oxide (ZnO), composed of zinc 
80.344 per cent, oxygen 19.656 per cent, specific gravity 5.42, corre- 
sponding volume 14.9, are changed to equal 125 parts of the carbon- 
ate of zinc (ZnCOj), composed of zinc 52.153 per cent, oxygen 
38.278 per cent, carbon 9.569 per cent, specific gravity 4.44, volume 
28.1. 

This chemical change, attended by so large an increase in 
volume (nearly double), if it takes place after the oxide has been 
spread as a paint, the whole coating will be loosened, and the loose 
particles will be carried away by storms; the action being similar 
to that which would occur should the sand mixed in the mortar of 
a plastered wall, when dried, change its volume to the same degree. 

Zinc oxide is therefore not a permanent paint for open-air ex- 
posures, but for interior use is permanent; for though carbonic acid 
is present, the moisture is absent, both elements being essential 
for the change to a carbonate. Dry, gaseous carbonic acid does not 
aflFect dry zinc oxide. It is mixed with white lead for interior use 
to lessen the tendency of the white lead to darken by absorbing the 
sulphuretted hydrogen present in nearly all locations, a small 
amount of which darkens the lead pigment. Such coatings are 
harder than the lead coatings, as the surface has changed in the 
exposure of drying to a carbonate of zinc. 



94 ZINC SULPHIDE AND ZINC SULPHATE. 

Zinc oxide is a hazardous pigment to use for external exposures 
when mixed with iron oxide, lead, or other color pigments. No 
process of mixing them can so associate the several pigments, even 
when ground in the oil, as to enable any one particle of either substance 
to thoroughly protect any particle of the others present, from the 
changes mentioned. Atmospheric moisture and gases will sooner 
or later reach the zinc oxide, whether in the first or other coats of 
the paint, and the inexorable laws of chemical change will govern the 
durability of the coating, the ultimate decomposition of which will 
be determined simply by the amount of zinc oxide present. 

Zinc sulphate (ZnSOJ is a white pigment, and is often produced 
in the manufacture of zinc oxide, the color of which is not affected 
by its presence, even if the sulphate is added to the oxide afterward. 
Zinc sulphate has recently been brought forward as a desirable pig- 
ment for ferric as well as for wooden surfaces. In Germany and 
England it is used largely in all mixed paints, and has thus far proved 
to be very resistant to atmospheric influences in damp locations. 
Like zinc oxide and the lead pigments, its cost is a great factor against 
its more extended use. The commercial zinc-sulphate paints are 
adulterated with barytes, the natural composition of which is favor- 
able for its admixture with the sulphate. 

Zinc sulphide is a pigment introduced by Mr. J. B. Orr, of Eng- 
land. Its composition is, barium sulphate 70.50 per cent and zinc sul- 
phide 29.50 per cent, the reactions being, BaS+ZnS04=BaS04+ZnS. 
The dense white precipitate formed is highly heated, then quenched 
in water, and finely ground and dried. Becton white. Oleum white, 
Orr's white, etc., are zinc-sulphide pigments. Great purity of the 
raw materials is required to produce a purely white product. 

Zinc sulphide is largely used in the manufacture of enamel paints, 
linoleum, table oilcloths, etc. It does not continue to oxidize 
after mixing with linseed-oil, as do lead oxides, and can be considered 
as a saturated or non-siccative compound. It does not combine 
with resin, and therefore will not saponify. Exposure affects it by 
blackening the paint if a lead drier has been used with the linseed- 
oil, or a lead pigment associated with the sulphide pigment. In 
these cases a lead sulphide is formed which is dark-colored. 

Lithopone or Lithophone is a German paint compound only 
lately manufactured in the United States. As a commercial mixed 
paint, it is composed of sulphate of zinc, zinc oxide, and barytes or 
Blanc-Fixe, generally, one-third of each substance, and is similar 



ZINC SULPHATE AND LITHOPONE. 95 

in character to Charlton's white and Griffith's patent zinc oxide- 
It is commercially classed as Green Seal, Red Seal, Blue Seal, and 
Yellow Seal. The Green Seal consists of one part zinc sulphate 
and two parts of barytes. The Red Seal, of one part zinc sulphate 
and three parts of barytes. The specific gravity of the Red Seal is 4.2, 
The Blue and Yellow Seals contain some zinc oxide with the sulphate, 
and a greater percentage of barytes, and are consequently deficient 
in covering power. The large amount of oil taken up by the sul- 
phate and zinc oxide is counteracted by the smaller quantity taken 
up by the barytes, 

Barytes costs about one-tenth as much as zinc oxide or zinc sul- 
phate, and affords every requisite to grade up the weight of zinc 
paints, even when a liberal amount of whiting is present, as is too fre- 
quently the case with most mixed pastes or color paints. (See Tests of 
Paints.) 

Greennseal lithopone approaches closely the best brands of French 
zinc oxide, and does not require so large an amount of thinners as the 
American brands of zinc white, and it works easier. It is unaffected 
by sulphurous gases, and does not turn yellow when thinned. It 
will blacken if exposed to the sun before it is dry. Oils or driers 
containing lead or copper salts turn lithopone gray; neither can it 
be used with other colors having a lead or copper base. 

Griffith's zinc white, a chloride or sulphate of zinc, is precipitated 
from a soluble sulphide or chloride of sodium, barium, or calcium. 
No iron should be present. The precipitate is dried, and then cal- 
cined at a low or cherry heat, with careful stirring; raked out of the 
furnace and quenched in vats of cold water, then levigated and ground. 
It is an oxysulphide of zinc ; some sulphate of magnesia accompanies 
the pigment. 

A commercial zinc white that is only a sulphate of zinc is made 
by precipitating the pigment, by the addition of a dilute solution of 
sulphuric acid, to an acetic- or nitric-acid solution of litharge. Wash 
and dry the precipitate thoroughly. The clear liquor can be used 
repeatedly. All of the metals associated with the zinc in the litharge 
are dissolved by the nitric or acetic acids and precipitated with them 
as sulphates, and are inclined to blacken on atmospheric exposure. 
The zinc pigments formed by the precipitation processes are not as 
durable or reliable as those formed by the oxidation or sublimation 
processes from metallic zinc or the zinc ores. 

A mixture of one part of zinc oxide with two parts of red lead 



96 ADULTERATIONS OF ZINC OXIDE, AND TESTS. 

has given very satisfactory results in retarding marine corrosion in 
both salt and fresh water.* 

The United States Bureau of C!onstruction specifies one part of 
white lead to three parts of zinc oxide for the paint used on wooden 
structures on the seacoast, and has lately abandoned the use of 
zinc-oxide pigments on ferric structures wherever located. House- 
painters use from 20 to 50 per cent of zinc oxide when they mix 
their own colors. The great percentage of zinc oxide in the com- 
mercial mixed-white paints and colors has been referred to in Chap- 
ter VI. 

Adulterations of Zinc Oxide, and Tests. 

Patent zinc white is a sulphide of zinc mixed with baryta or 
strontia. Fulton's zinc white is the sulphide of zinc and barytes. 
Charlton's zinc white is the same. 

In the adulterations of zinc oxide with barjrta, bar)rtic white, 
permanent white, Blanc-Fixe, constant white, etc., all of these sub- 
stances are artificial sulphates of baryta, and are less crystalline than 
the natural sulphate, and cover better. Pure zinc oxide dissolves 
entirely in dilute sulphuric acid, leaving no residue. If carbonate 
of lime is present, it effervesces with muriatic acid, and the amount of 
this action in a measure indicates the amount of that adulterant 
present. 

Zinc oxide lacks weight when compared with white lead for paint 
mixtures. Barytes in its powdered form supplies this deficiency, 
but has poor covering power; it spreads well and saves oil. The 
floated barytes — a finer grade floated from the pulverized natural 
mineral — has better covering power than the ordinary brands of this 
pigment, simply because it is finer. Artificial barytes or "Blanc- 
Fixe" frequently contains pulverized silica or white-glass sand. All 
of these substances are adulterants, and add nothing to the qualities 
of zinc oxide except weight and a saving of oil that lessens the cost 
of the mixture to the manufacturer, but seldom to the consumer. 

To test a mixed zinc-oxide paste or paint for adulterations, repeated 
washings with benzine or ether will remove the oil; then dry the 
residuum on blotting-paper. Dilute sulphuric acid completely dis- 
solves zinc oxide, leaving the adulteration or any other metallic-base 
pigments unaffected. 

♦ " United States Navy Yard Tests of Marine Paints," Transactions 
American Society Mechanical En^neers, Vol. XVI, 1894. Paper Number 
625, p. 390. 



ADULTERATIONS OF ZINC OXIDE, AND TESTS. 97 

A dilute solution of muriatic acid will dissolve the lime, if any is 
present. The barytes, silica, etc., will remain after the residuum 
has been ignited. 

On a pamted surface, slightly scratch the coating, and apply a 
drop of sodium sulphide of 100® Baum4. If lead pigments are present 
a discoloration will follow the application of the sodium. 



CHAPTER VIII. 

LAMPBI^CK. 

The carbon group of pigments comprises lampblack, mineral 
or natural asphaltum, artificial asphalt, coal-tar, and graphite, either 
alone or as a component part of the paint. 

Lampblack is the fine deposit or soot formed by the imperfect 
combustion of oil or fatty substances. Its composition varies greatly, 
depending upon the nature of the substance consumed in its forma- 
tion, and the care exercised in the combustion process. Fatty oils 
and grease yield the best lampblack. Coal-tar yields a black of a 
brownish hue and is inclined to be oily. Resin furnishes a good black. 
If the combustion is forced it carries along some of the free resin- 
flakes, and yields a yellowish resinous black of an inferior quality, 
not always free from grit and dirt. 

Gas-blacky the soot product from the combustion of hydrocarbon 
fuel or illuminating coal-gas, differs in molecular structure from the 
fatty-oil blacks. The gas-black particles appear to have a star form, 
and are not as suitable for mixing with white lead or zinc white for 
tints as the fatty-oil blacks, though their color is densely opaque. 
The fatty-oil lampblack is filament-formed, and incorporates with 
the oil and oxide pigments better than the star or flake-formed blacks. 
Gas-black is also made from natural gas burned under revolving 
cylinders, the deposited soot being removed by scraping. With 
proper care the lampblack so formed is nearly pure carbon. In a 
paint coating it has a tendency to become brittle, crack, and flake 
off after a short time. This possibly results from using too much 
drier or turpentine in the vehicle, as of itself it is a slow-drying 
pigment, and adds no drying qualities to the vehicle. It seldom 
appears in the market as gas-black. 

Ground soot appears as a lampblack under various trade-names. 
It contains ammonia, sulphuric and pyroligneous acids, rain- 
water, and carbonic acid. Under atmospheric conditions, solutions 

of these acids are produced strong enough to set up galvanic action 

98 



LAMPBLACK. (CARBON GROUP OF PIGMENTS.) 99 

on roofing material and ferric surfaces. With galvanized iron or 
sheet zinc, the zinc is reduced to either an oxide, sulphate, or carbon- 
ate at the expense of the zinc covering, leaving the iron exposed 
to the action of the elements which produce corrosion, that is the 
more active because of the galvanic couple of the different metals. 
(See Galvanizing, Chapter XVII.) 

Spanish black, or cork-black, is made from the combustion gases 
of burning cork. It is a good lampblack in color and texture if proper 
care be taken in the process; but charred cork and ashes are too often 
present in the product for its good. 

Ivory-black is made from chips of elephants' tusks and other 
hard bones free from fat. It should have no lustre, as that 
indicates the presence of unconsumed fatty matter. Its use is 
almost exclusively for the preparation of the finest blacks for carriage, 
decorative, and artists' colors. Its high cost debars it for use in 
ferric coatings. 

HorrirbUicky or animal black, is almost identical with bone- 
black, but is generally in a more finely divided state. Ani- 
mal refuse, albumen, gelatine, horn and hoof shavings, etc:, are 
subjected to a dry distillation in a still or retort; the black carbo- 
naceous mass left is washed with water and powdered in a mill. It 
requires about one and a quarter its own weight of oil for a paste. 
The great quantity of oil left in the black as it comes from the still 
is the reason for its slow drying. It is a cold, mild black, and when 
not well burned has a brown tint, dries badly, and is used for printers' 
ink, blacking, etc., also for the cheaper grade of black varnishes and 
paints. 

Bone-black, made from a poorer quality of bones than ivory-black, 
is a warm, reddish-brown black. 

Drop-black is an ivory- or bone-black blued with Prussian blue. 

Charcoal-black is a finely powdered beechwood charcoal, made 
in Sweden, and generally marketed as Swedish black. It is a pure 
black in color, but has less covering power than the fatty-oil blacks. 

BliLc-black, made from vine-stems, is a better quality of charcoal- 
black. 

Frankfort black, or vine-black, is made from the charcoal left 
from the calcination of dried vine twigs, wine lees, peach-stones, 
bone and horn shavings, etc., and contains potash. It varies in 
shade according as the animal or vegetable charcoal is in excess. 
The animal matter gives it a brownish hue, the vegetable a bluish 



100 LAMPBLACK. (CARBON GROUP OF PIGMENTS.) 

color; both have a good covering power. The finest qualities of 
this black come from the condensed gases as soot in the calcination 
of the above substances. Many other process blacks are sold under 
the above names. 

Almond-4)lack is made from fruit stones, nuts, etc. It is an intense 
blacky and has the same qualities as Frankfort black. 

German black is made from the combustion gases of any resinous 
matter, which escape into a long flue, at the end of which is a woollen 
or other fibre hood, that collects the deposited soot. 

English black is collected from the flues of bituminous coke-ovens. 

RiLSsian black is made from the soot of resinous dead fir- or pine- 
wood. It is liable to spontaneous combustion if left for a short time 
moistened with oil. 

Prussian black, Berlin black, ochre-black, coffee-black, earth-black, 
lake-black, paper-black, and manganese-black are all inferior qualities 
of lampblack made by some one of the many processes, and from 
the many substances capable of slow and imperfect or smoldering 
combustion. Their color and qualities are quite as divergent as 
their names; all dry slowly with uncertain results in color and 
lustre. 

Graphite-black, or ship's black, is an impure lampblack mixed 
with an inferior quality of flake-graphite, and can be known by its 
metallic lustre. 

Coal- or shale-blacks are generally pulverized slaty bituminous 
coal. 

The trade adulterants of lampblack consist not only of those 
substances that in the process of manufacture are imperfectly car- 
bonized or vaporized, but nearly every other light substance that 
is black and can be ground to the required fineness. The coarse soot 
and scales deposited in the chimneys and flues from the combustion 
of fatty-wood and soft-coal fires, coal-gas, mineral oil, shale, and 
asphaltum, coal-tar, etc., in the several processes of distillation or 
burning for other products, all contain ashes; also acetic, pyroligneous, 
and sulphuric acids, ammonia, and tar to a notable extent, that con- 
dense in the carbon-atoms, and materially affect the color and quality 
of the lampblack. These are not always removed in the subsequent 
calcination that all lampblacks require to form the prime pure 
product known as " burnt lampblack." These acids and the tar pre- 
vent the drying of any lampblack coating, except by the use of an 
excessive amoimt of strong driers. In such cases the paint hardens 



LAMPBLACK. (CARBON GROUP OF PIGMENTS.) 101 

only on the surface, and remains viscid underneath, and is prone to 
peel. 

Anthracite and bituminous coals are ground and marketed as 
pure lampblack. They contain from 8 to 12 per cent of ash, also 
from J to 2 per cent of sulphur, and absolutely have not a single 
quality to recommend their use except their low price. From the 
large quantity of worthless lampblack selected for the finishing 
coating of most of the ironwork in the New York Rapid Transit 
Subway, it might well receive a special trade-name as the "Subway 
or Tunnel black." 

Carbon-bldck appears in the market as hydrocarbon-black, Ameri- 
can gas-black, satin-black, gloss-black, jet-black, silicate of carbon, etc. 

To make a pure lampblack requires not only a proper material, 
but as careful attention to the combustion of it and the subsequent 
processes for its preparation as the manufacture of any other pig- 
ment. 

Pure lampblack made from a fatty oil is so finely subdivided 
naturally, that it requires no grinding. It is only ground in the 
vehicle to secure a more thorough incorporation than is possible 
by stirring it in. It is of an oily feel and nature, and in combination 
with a good oil forms a more elastic and closer-clinging coating than 
any other pigment. 

It is chemically and electrically passive, non-hygroscopic, non- 
corrosive, and less affected by heat, light, and peeling than any other 
pigment. 

Its life in a paint coating is in a great measure exempt from all at- 
mospheric influences that cause the decay of a paint. Its elastic 
nature reduces the frictional element due to the beating of storms, 
while the oxidation or decomposition of organic matter in the dust 
and from other sources is almost nU, It remains in place until re- 
moved by friction or the destruction of the vehicle, and can be 
painted over without the expensive torch-burning or scraping so 
necessary with other pigment coatings. 

In some form or degree of purity lampblack enters into all of 
the black varnishes, enamels, and trade paints that have any marked 
quality for the protection of metallic or other surfaces. From its 
finely divided state and oily nature it is liable to spontaneous com- 
bustion, hence must be stored in small bulk and kept well covered, 
lampblack requires more than double its own weight of oil to secure 
a good coating; it is easily brushed out with but small wear of the 



102 LAMPBLACK, (CARBON GROUP OF PIGMENTS.) 

brushes. Driers added to lampblack paste or varnishes should be in 
the form of japans, rather than turpentine, which flattens the lustre 
of the coating. 

Many instances are on record where a single coat of lampblack, 
like that used for the lettering and symbols on the old cross-road and 
tavern sign-boards, that have been exposed for a century or more, 
are still uninjured, while the surrounding colors and in many cases the 
wooden surface of the sign have been worn away, leaving the carbon 
lettering in full relief. 

The iron-link-chain suspension bridge over the Merrimac River 
at Newburyport, Mass., was made of Norway cold-blast iron, and 
erected in 1810. It was painted with two coats of pure lampblack 
and raw linseed-oil over sixty years ago, and is still (1903) practically 
free from corrosion, though in an exposed position, subject to sea air 
and fog influences for days in succession. 

The use of lampblack to delay the "setting*' of red lead is fully 
described in the article on red lead. It does not, however, prevent 
the failure of red-lead coatings when they are exposed to the action 
of hydric-sulphide fumes, but is not itself affected by them. 

A good test of the quality of a lampblack is to place the sample on 
a piece of blotting-paper and pour a little ether on it mitil the paper 
is soaked with the ether, percolating through the black. If on the 
evaporation of the volatile and removal of the powder the under side 
of the paper appears fatty, the lampblack is of poor quality. 

Animal charcoal and bone-black or ivory-black are strong bleaching 
agents, and it is possible for them to uncolor overlying coatings. The 
oil protects them somewhat from this bleaching influence, but where 
long stability of color or lustre is required, it is better to use blacks not 
of an animal nature. 



CHAPTER IX. 

MINERAL OR NATURAL ASPHALTUM. — ^ARTIFICIAL ASPHALT (WHICH 
INCLUDES COAL-TAR AND ITS PRODUCTS, PITCH, MINERAL WAXES, 
ETC.) . 

Mineral or natural aaphaUurn. There are a large number of these, 
known as Egyptian, Bermudez, Trinidad, Mexican, Cuban, Cali- 
fomian, Chinese, etc. They all vary greatly in character and purity, 
and are the residual products of petroleum when the light hydro- 
carbon elements have been evaporated by natural causes. They 
contain vegetable and mineral matter, sulphuric and other acids 
that must be removed by boiling or distillation to render them suita- 
ble for enamels, varnishes, or paints. Asphaltum is not to be con- 
founded with the product of coal-tar distillation, called "asphalt,'^ 
which, having a certain resemblance to the natural asphaltum in 
some of its physical qualities, is chemically very unlike it. The name 
asphalt being carelessly applied to both the natural and artificial or 
coal-tar product, naturally leads to ^ome confusion on the subject. 
They are, in fact, so widely apart in all their essential qualities that 
they cannot be appropriately coupled^ together as relating to the 
same substance. 

The characteristics of asphaltum used for ferric coatings are 
briefly given: Asphalt, bitumen, or mineral pitch, specific gravity 
1 to 1.68, softens at 170° F. and melts at 212** F. (coal-tar asphalt 
softens at 115° F.). According to Boussingault (Am. Ch, Phys. [2], 
XIV, 141) it is a mixture of two definite substances^ viz.: aaphaUene, 
which is fixed and soluble in alcohol; and petrolene, which is oily and 
volatile. The greater part of the latter may be volatilized by dis- 
tilling the asphalt with water. The chemical composition of bitumen 

is: 

Carbon 85 per cent. 

Hydrogen.. ..12 " " 
Oxygen 3 " " 

100 " " 

It is therefore an oxygenated hydrocarburet. 

103 



104 MINERAL ASPHALTUM. 

It is the petrolene that gives the cementitious or bonding value to 
compositions into which it enters. Bermudez asphalt is about 2 to 3 
per cent purer than Trinidad. Samples of Bermudez analyze 97.22 
per cent of materials soluble in bisulphide of carbon. A large amount 
of these materials is also soluble in ether, showing that the bitumen 
contains large amounts of petrolene. 

Petrolene in Bermudez = 81.63. 
" Trinidad =80.01. 

Egyptian asphalt is the purest of all the varieties of asphalt, but 
is not procurable at present in commercial quantities required for 
pavements or paints, but is used in the finer qualities of japanned 
or enamelled wares, baked coatings, varnishes, etc. Samples of it 
frequently analyze 99.5 per cent of soluble matter. 

Asphaltum yields by dry distillation a yellow oil, consisting of 
hydrocarbons mixed with a small quantity of oxidized matter. It 
begins to boil at 90° C, but gradually rises to 250° C, giving oils of 
specific gravity during the boiling, viz. from 90° C. to 200° C, specific 
gravity = 0.817 (at 15° C); that which boils between 200° C. and 250° 
C, specific gravity = 0.868 (at 15° C.) ; both portions giving by analy- 
sis 87.5 carbon, 11.6 hydrogen, and 0.9 oxygen, which is nearly the 
composition of the oil of amber. 

These asphaltum oils, treated with sulphuric acid and then washed 
with p)otash and subjected to dry distillation, yield a number of oils 
which are insoluble in water, or strong nitric acid, and are but little 
affected by strong sulphuric acid, but are very soluble in alcohol or 
ether. 

Asphaltum has no metallic base, and can be classed as a gum or 
resin, hence but a small amount of it can be incorporated into an oil 
vehicle for use as a paint. Bisulphide of carbon and benzine usually 
form a large percentage of all vehicles in asphaltum paints. 

The principal merit of some of these paints consists more in the 
name than the quality. If it is once considered that only about 10 
per cent of asphaltum enters into the composition of the well-known 
street pavements, and that so little quantity as this amount, however 
it may govern the other constituents of the paving compound, has to 
be put in place or applied hot, and cannot be used or compounded in 
any other manner, it may be apparent that, notwithstanding the 
catchpenny name, really but little if any asphaltum of either high or 
low degree ever enters into the composition of any of these paints. 



MINERAL ASPHALTUM, 



105 



Analyses of many of these paints show that there is not 5 per cent 
of asphaltum in the composition of any brand of such paint upon the 
market. Even with this small amount, and with the best of boiled 
or raw linseed-oil as the vehicle, the paint is difficult to dry without 
the use of strong metallic salts mixed with the oil to aid its oxidizing 
or drying quality ; and if a quick-drying paint is wanted, these oxidiz- 
ing materials are added in such amounts as to materially affect the 
life of the paint. 

When the color of the paint is other than black or steely gray, it 
may be doubted if any asphalt will be found present under the closest 
analysis; and the red and brown colored samples will be found to 
rely almost wholly upon oxide of iron as the base of the pigment, 
under whatever name it may be masked. 

Gilsonite, a mineral resin associated with natural asphaltum, is 
used largely as the principal pigment in these paints. Gilsonite, 
asphaltum, petroleum, cannel and bituminous coal and shale, all shade 
off into each other so gradually, and form so numerous a class of bitu- 
minous mineral substances, that it is difficult to determine their exact 
relations. The fluid elements of the hydrocarbons evaporate, and as 
the heavier portions solidify, they oxidize with a loss of hydrogen, 
and change until over a hundred different bituminous mineral sub- 
stances can be determined from the hydrocarbon group. 

The general composition of the numerous class of petroleums, 
after the evaporation of the lightest hydrocarbons by nature in the 
form of natural gas, is, viz.: 



Grade Oil 26** Baum^. Distillates. 
Commerrial Names. 



Gasoline 

Benzine 

Kerosene (illuminating-oil) 

Heavy kerosene (mineral sperm) 

Gas distillate 

Light lubricating (spindle-oil) 

Neutral oil 

Heavy lubricating-oil 

Valve lubricating-oil 

Asphalt (crude) , containing 4 to 7 per ) 

cent of sulphur j" 

Ix)ss 



Approx- 
imate 
Degree. 
Baumd. 



75-76 

63 

45 
38-40 

28 

26 

23 

21 
14-15 

11-6 



Specific 
Gravity. 



0.6820 

. 7253 

0.80 

0.8333 

0.8866 

0.8974 

0.915 

0.9271 

0.9655 

1 to 1.60 



Weight of 1 

U. S Gall, in 

Pounds. 



5.69 
6.04 
6.66 
6.94 
7.38 
7.48 
7.62 
7.72 
8.04 
8.344 to 
13.350 



Percent- 
age Ob- 
tained. 
Approx- 
imate. 



3-5 
4-6 

13-15 
8-56 

10-18 
8-10 

10-12 
5-6 
4-5 

11-12 

5 to 13 



Other samples of petroleum range from 5° to 6° Baum^ higher, 
and carry more hydrocarbons of the paraffin series. 



106 ARTIFICIAL ASPHALT AND COAL-TAR. 

The illuminating parts of these oils carry more carbon, and less 
hydrogen, and give a smoky flame, due to the fact that it requires 
more oxygen to effect complete combustion of the carbon element 
than it does to consume the hydrogen. 

Coal-tar is a generic term applied to those bitumens which are 
extracted during the destructive distillation of bituminous coal for 
gas or coke. Commercially, the name is also applied to water-gas 
tar. The nature of the tar varies with the nature of the coal, and 
with the processes employed in its* production as a waste product in 
the manufacture of gas or hard coke. 

There is no known method of describing accurately the true com- 
position of the coal-tars. No two are identical in every respect, 
although many are identical in every essential respect. Variations 
also occur from the admixture with the coal in process of distillation, 
of greater or less quantities of oils of various kinds, used for the pur- 
pose of enriching the gas. The tars vary in the amount of bitumen 
they contain within the limits of 60 to 92 per cent; also vary largely 
in the percentages of oil which they contain, and in the quality of the 
oil. The non-bituminous matter in the tar is generally carbon, 
which is sjmonymous with lampblack, and was, of course, a hydro- 
carbon before the hydrogen was eUminated by combustion. Coal- 
tar cement, or asphalt, is a residue from the distillation of coal-tar. 
Its hardness or flexibility is due to the percentage of the oil left in it, 
and may vary from 16 per cent in one quality of coal-tar to 52 per 
cent in another. One per cent of oil taken from one coal-tar will 
produce a greater hardening effect than IJ per cent taken from 
another tar, and the degree of heat necessary for distilling off the 
oil may vary from 200° to 600® F., even when supplemented by 
mechanical agitation, or by blowing superheated steam or air into 
the still during the distillation process. 

The average analyses of a large number of samples of coal-tar 
from coal-gas retorts gave for a 40-gallon barrel, specific gravity 1.08- 
1.10: 

1} gals., or 3.75%, of light oils, consisting of benzole, naphtha, and carbolic acid, 
91 gals., or 23.75%, of heavy oil, consisting of creosote-oil and anthracine, etc. 
29 gals., or 72.5%, pitch. 

Boiled in open kettles, this tar should be reduced trom 15 per 
cent to 25 per cent, according to the duty required of it. The tar 
resulting from the distillation of petroleum oils for water-gas is of 
a decidedly inferior quality to that obtained from gas-coals, and is 



ARTIFICIAL ASPHALT AND COAL-TAR. 107 

better adapted for coating the cruder forms of wood constructions, 
piles, dock-timbers, fence-posts in the groimd part, than metal-work. 
But this same oil-tar, if distilled at heats from 600° to 800° F., 
fonns a pitch of almost adamantine hardness when cold, and resists 
almost all corrosive agents and solvents except those of the hydro- 
carbon class. 

Analysis of a by-product coke-oven tar: 

Naphthalene 12. 00 per cent 

Anthracine 0. 30 " 

Tar acids 7.00 " 

Tar bases 1.60 " 

Water 2.00 " 

Pitch 77.10 " 



100.00 " " 

When the concentration of the gas coal-tar is carried to the 25 
per cent or 30 per cent stage, the product is comparatively odorless, 
or at least is not any more objectionable than that from oil paint, 
the pungency due to the light oils and carbolic acid being dissipated. 
In the distillation of coal-tar, until the final residuum of coke 
is reached in the still, there are no constituent oils derived from 
the process that do not gradually volatilize by the heat of the sun 
or approximating temperatures; and all coal-tar or hydrocarbon 
products suitable for use in, or as paints, also become fluid when 
exposed to heat; in fact, but few of them are applied in any other 
condition than while hot. They are all liable to run on vertical 
or slightly inclined surfaces, until by evaporation they are so ad- 
vanced on the road to brittleness that they solidify, and by a little 
further progress in the same direction they become brittle and scale 
ofif on the least mechanical disturbance. 

In the production of an ordinary standard roofing-pitch from a 
coke-oven tar, the distillation ran, viz.: 

Water 1 . 49 per cent 

Light oil (to 325° F.) 3.04 " " 

Heavy oil (above 325° F.) 21 . 47 " " 

Pitch (at 585° F.) 74.00 " " 



100.00 



<( It 



108 ARTIFICIAL ASPHALT AND COAL-TAR. 

The pitch contained about 12 per cent of free carbon. Coal-tar 
asphalt softens at 115° F. Natural or mineral asphalt softens at 
150° to 170° F. 

Analysis of a standard coal-gas tar (specific gravity 1.24): 

Carbon 89 . 21 per cent 

Hydrogen 4. 95 " '' 

Nitrogen 1.05 " " 

Oxygen 4.23 " " 

Ash trace 

Volatile sulphur 0. 56 " " 

100.00 " " 

In combustion it gave British thermal units 15.781. Evaporative 
power from and at 212° P\ = 16.4 pounds of water. 

Analysis of water-gas tar from gas-oil (specific gravity 1.15): 

Carbon 92. 70 per cent 

Hydrogen 6. 13 " " 

Nitrogen 0.11 '' " 

Oxygen 0. 69 '' " 

Ash trace 

Volatile sulphur 0. 37 " " 

100.00 " " 

In combustion it gave British thermal units 17.193. Evaporative 
p^)wer from and at 212° F. = 17.8 pounds of water. 

In the distillation of an oil water-gas tar by Dr. John F. Wing, 
the products obtained at the several stages of the process were as 
follows (specific gravity of the crude tar 13.5° Baum^ = 1.1 — water 1) : 

Distillation heat, F. Perc(?ntage of distillate. 

240° F. water 0. 25 per cent 

240° F. light oil 4.25 " " 

240° to 336° 0.50 " " 

336° " 400° 3. 00 " " 

400° to 550° 29 00 " " 

550° *' 617° 5.00 '* " 

617° " 690° 17.00 '' " 

Above 700° a hard-pitch residue 41 . 00 



t( (t 



100.00 



tt tl 



ARTIFICIAL ASPHALT AND COAL-TAR. 109 

At 400° F. the distillate became heavier than water. The residues 
obtained at temperatures of 550® to 617*^ F. were soft pitch, but would 
not flow. There were from 12 to 15 per cent of free carbon in the oil- 
gas tar, while 5 to 8 per cent are the usual amounts found in an Otto- 
Hoffman coke-oven coal-tar. 

^ The acids and ammonia salts in crude coal-tar must be eliminated 
by boiling or distillation when used for coating ferric bodies. If 
they are not removed, the tar, either hot or cold, is one of the most 
unreliable and unmanageable of coatings. (See Dr. Angus Smith's 
experience. Chapter XII.) 

The mineral waxes derived from coal-tar are the most reliable 
of all the coal-tar paint products. They are especially not affected 
by "sweating." They are an intermediary substance between the 
fluid and volatile elements and the heavy ones; and retain some of 
the volatile element that, as it slowly evaporates, causes the paraffin 
to crack badly and change its volume. The spaces between the 
tension-chord and other eye-bars in modem bridge constructions, lying 
so closely together as to be incapable of inspection or repainting to 
protect them from corrosion, are often filled in with melted paraffin 
as a protection from rust. It requires but a short period for the wax 
to harden, shrink, and crack, and expose the ferric bars. As well 
expect a cracked varnish coating to protect the surface it covers, 
as one of cracked paraffin. 



CHAPTER X. 

ASPHALTUM PAINTS AND CARBON VARNIS^ES. 

AsPHALTUM paints are proprietary products, and vary in composi- 
tion and quality quite as much as does the substance from which they 
derive their name. There is no standard of excellence in asphaltum 
paints. 

A small amount of some quality of mineral asphaltum or gilsonite, 
mixed with varying amoimts and qualities of the trade lampblacks, 
constitutes the pigment for the numerous brands of quick-drying 
paints used to blacken a large class of ferric bodies that need a 
coating for appearance rather than protection from corrosion. The 
catchy name often secures their use on more important structures, 
where the price at which they are offered should promptly condemn 
them before trial. 

A supposed better class of asphaltum paints or so-called var- 
nishes, similar to the "Maltha," "P. & B.," and other trade-mark 
designations, are freely marketed as superior paint products. They 
are in no sense varnishes, but simply the above-mentioned class of 
pigment substances, mixed with bisulphide of carbon, benzine, and 
other uncertain hydrocarbon liquids and oils, the latter often con- 
taining more resin-oil than linseed-oil. They are not compounded 
by heat, as all true varnishes are. They have had an extended trial 
for over fifty years on important ferric structures, — naval, hydraulic, 
and other work, only to fail after a brief exposure. Wherever placed 
in competition with other carbon or metallic-base coatings they are 
invariably found low in the column of merit. As a rule they spread 
ea^ly and show well at first, but when the volatiles have evaporated, 
especially if they have been subjected to a moderate heat test 140® to 
180® F., they become brittle, turn brown, crumble, and are easily 
removed. The application of these paints, containing bisulphide 
of carbon, is attended with extreme danger from fire, even on external 
exposures. The vapor of bisulphide is very explosive at low tem- 
peratures, also disastrously injurious to the painters or others breath- 
ing it during the application of the paint in any confined space, and 

only moderately less so in the open air. 

110 



Flo. 20.— Animi Fossil Res 



FOaaiL RESINS. Ill 

Alt account of its application to water^mMns, where it resulted in 
the insanity and death of a number of the painters and woitmen 
engaged in painting and laying the pipes; also in the utter failure of 
the coating to protect the same pipes from corrosion, is given in " Trans- 
actions American Society Mechanical Engineere," Vol XVI, 1895, 
Paper 637. Also in Engineering News, Feb. 7, 1895, and April 4, 1895. 

A further demonstration of the inferiority of these asphaltum 
paints in competition with other oil painta and black vamishea ia 
given in a series of tests made by Mr. Max Toltz, C.B. The Report 
was read before the Society of CivU Engineers, St. Paul, Minn., and 
n^rted in the Jmimal of the AssocuUwn of Engineering Sodeliea, 
\mt. It was also briefly referred to in " Transactions American 
Society Mechanical Engineers," May, 1901. (See also Bisulphide of 
Carbon, Chapter XX.) 

Asphaltum varnishes or carbon pfunte in which the vehicle is 
practically a Unseed-oil varnish, compounded by heat, and of the same 
nature as a baked-japan vehicle in which the carbon-blacks and 
other pigments are ground, are very reliable for protective coatings. 
They seldom fjul under the severest tests of marine or other corrosive 
exposures. 

Fossil Resins. 



Fia. 19. — Section through a reBin paasage of Abits exceha (fir- and spruce-tren). 
The cavity Hg, aa well as the thin-walled cells Hp, are filled with semi- 
fluid reain. llie thick-walled cells P contain etarclL 

Fossil gums or resins, under the general name of Copals, are those 
used for varnishes or varnish paints. They are incorporated by 
heat with refined linseed-oil, and when black generally contiun a 
quantity of the better class of refined asphaltum. 

The oldest and hardest of the fossil reeins is Zanzibar; the trees 
that furnished it are extinct. 



112 FOSSIL RESINS. 

There are about thirty dififerent resins used in paint and varnish 
manufacture, many of them possessing peculiar qualities. The best 
are the fossil resins found in the beds of rivers or in the earth where 
they have lain for centuries. The hardness of these fossil gums appears 
to depend upon their age and the pressure that they have undergone 
while buried. Amber is the hardest and most valuable of all resins. 
Only the refuse of black amber is used for varnish. Amber varnish 
merely means amber-colored varnish. There is no amber in the com- 
mercial brands. 

Copal is the next in hardness; it comes from Zanzibar, and is 
known in the English trade as "Animi," from the insects embedded 
in it. Being very difficult to dissolve, it is distilled until it loses 
from 20 to 25 per cent, when it can be dissolved in boiling oil. There 
are three varieties of it, and many grades. 

*'Animi" is now the technical name for the South- American 
copal, and comes from Brazil. 

Sierra Leone copal has nothing to do with Sierra Leone except for 
its name. It comes from the river-beds in the interior of Africa. 
It is the only African resin that will dissolve in cold alcohol. Its 
color is not as good as the Zanzibar or best Kauri, but it is harder 
than the Kauri. It is mixed with the Zanzibar for hardness, itself 
giving toughness to other fossil resins. 

Other African copals are the Pebble or Pebble-stone — ^which is the 
hardest — Acora, I^oango, Gaboon, Congo, Benguela, and three sorts 
of Angola. 

Manila is of two kinds, — a hard and a soft; neither are fossil 
gums. They come from the Philippine and other islands, Borneo, 
Singapore, etc. This gum can be used as it comes from the living 
tree like the crude resin from the American long-leaf pine. 

Dammiir is a recent resin from trees not extinct, and contains 
the most water. When it forms the principal resin in a varnish or 
varnish paint, it appears to be always drying, hence the danger to 
any other coating spread over it. It is the resin used with enamel 
paints to give the high gloss characteristic of these coatings. 

Sandiira^ch is a resin yielded by the barberry-trees of Northern 
Africa. It is used to a considerable extent as the basis of spirit var- 
nishes. 

Kauri or Cowrie, from New Zealand, is the principal fossil resin 
used for a varnish, being about ten times the amount of all the other 
resins combined. It is produced from a species of tree not yet extinct, 



FOSSIL RESINS. 113 

but the gum as it exudes from the tree at the present day is of no 
more value for a varnish than that from the common spruce-tree; but 
when it has lain in the earth for centuries it becomes hard and valu- 
able. It is veiy indifferent to the action of sulphur gases, and is 
more colorless than the other fossil resins. It is easilj*^ dissolved, 
and melts more readily than mastic, but less so than the common 
resins. It is allied in composition to Dammar resin, and is from 
two to nine times cheaper than the other fossil resins whose prices 
range in the order of conmiercial quantities as follows: 

Prices per pound. 

First. Kauri 10 to 50 cents. 

Second. Manila 10 " 25 " 

Third. Dammar 16 " 25 " 

Fourth. Zanzibar, best $1 " $1 .25 

Fifth. Benguela 85 " 90 cents. 

The general composition of all fossil or other resins is CjoH^oOj. 
Their specific gravities at 60° F. are: 

• 

Yellow-leaf pine-resin (dark colophony). . . . 1.100 

" " " " (whitish opaque) 1.047 to 1.044 

" " " " (yeDow transparent). . 1.084 " 1.083 

Shellac, D. C. (dark colored) 1. 123 

" L. C. (light colored) 1. 114 " 1.113 

" B. (bleached) 0. 968 " 0.965 

Copal, East Indian 1.070 '' 1.063 

" West Indian 1.080 " 1.070 

" Very old 1.055 " 1.054 

'' Zanzibar 1 .068 " 1.067 

Dammar, Manila 1. 121 " 1.062 

old 1.075 

Benzoin, Siam 1 .235 

" Penang 1.155 " 1.146 

" Borneo 1.170 " 1.165 

Guaiacum, pure 1.237 " 1.236 

Tolii, old and brittle 1.232 " 1.231 

Amber 1.094 " 1.074 

Sandarach 1.044 " 1.038 

Mastic 1.060" 1.056 

Angola 1.081 " 1.064 

Kauri (Australian) 1. 115 " 1.050 

Brazilian 1.082 " 1.018 

Shellac is an animal resin produced from the banyan- or fig-tree 
and other trees of India, called "Lac-trees.*' Lac is the root of the 
word Ijacquer, of Indian derivation, and was probably first applied 



114 



FOSSIL RESINS AND SHELLAC. 



to specimens of Chinese lacquer-ware, imported through India. The 
branches of the lac-trees, when stung by the female insect Coccus- 
lacca, exudes a sap that the insect transforms by digestion into a 
resinous excretion (lac), with which she encrusts her eggs and herself. 
The insect is indigenous to the forests of India. The exudation of 
the sap from the lac-tree is somewhat similar to that produced by 
an insect or parasitic fungus on a species of oak (the gall-oak) that 
produces the "gall-nut" used by dyers and in pharmacy. The lac, 
when sent to market, often contains the eggs of the insect, and is 
called "seed-lac." 

The lac secretion is dissolved from the twigs and branches of the 
tree in hot water, the solution is then evaporated on hot revolving 
cylinders, or in shallow pans, then scraped off in the form of thin sheets, 
broken up, and forms the commercial shellac, graded generally as 
D. C. (dark colored), L. C. (light colored), B. (bleached), etc. 

The coarse qualities of the melted lac, when dropped into rounded 
pieces 1 to 1^ inches in diameter, are called "button-lac," and when 
in larger pieces are known as sheet-lac or "piece-lac". 

The best quality is kuaum-lac, from the kusum tree {SMeicheror 
trijnga), which lasts about ten years after being stung. The twigs 
from this tree are of a light-golden color and furnish the orange shellac ; 
coming principally from Siam. The second quality is furnished by 
the dhak or polos, from the Butea frondosa. The third quality is 
the pipal, from the Fiscus rdigiosa. All of the lac-trees except the 
kusum live only from two to three years after being stung by the 
insect. Commercial shellacs are extremely variable in quality and 
price. The best grade of fine Orange D. C. lac brings £10 125. per 
cwt. in London; "Native Orange," £8 9s. per cwt.; "Garnet," 
£7 Ss.; "Native leaf" and "Button," £3 Ss. to £3 6s. 

The composition of shellac is given by Mr. Halstead* as 



Resin 

Coloring-matter 

Wax 

Gluten 

Extraneous and nitrog- 
enous substances. . 
Loss 



Stick-lao. 



68.00 percent. 
10.00 " " 

6.00 " " 

6.60 " 



(( 



6.60 " 
4.00 " 



tt 



100.00 percent. 



Seed-Uc. 



Shellac. 



88 . 50 per cent. 
2.50 " " 
4.50 " " 
2.00 " " 



1.00 " 
1.60 " 



it 
tt 



100.00 per cent. 



90 . 90 per cent. 
0.50 " " 
4.00 " " 
0.00 " " 



2.80 " 
1.80 " 



tt 
tt 



100.00 percent. 



♦ Geo. Watts's Dictionary of the Economic Plants of India. Government 
Printing-office, Calcutta, ^889. 



FOSSIL RESINS AND SHELLAC, 116 

A more complete analysis by Dr. Johns* shows that 120 parts of 
stick-lac consist of 



An odorless common resin , 

A resin insoluble in ether 

Coloring-matter analogous to cochineal , 

Bitter balsamic matter 

Dun-colored extract , 

Acia (laccic acid) 

Fatty matter like wax 

Skins of the insect and coloring-matter (the latter furnishing 

food for the grub when hatched) 

Salts 

Earths 

Loss ^ •. 



80.00 parts 
20.00 

4.60 

3.00 

0.50 

0.75 

3.00 

2.50 
1.25 
0.76 
3.76 



120.00 " 



Shellac dissolves readily in alcohol, benzine, muriatic and acetic 
acids, but not in concentrated sulphuric acid. It dries solely by the 
evaporation of the solvent, leaving the thin film unchanged, the 
only use of the solvent being to spread the varnish. When alcohol 
is used as the solvent, the varnish can be spread over damp surfaces, 
as the alcohol will take up the moisture without much apparent 
injury to the coating, though this will be longer in drying, as the 
water must be evaporated with the alcohol. 

Shellac can be applied to ferric surfaces, and in under-water (fresh) 
exposures it generally will remain about two years without any great 
deterioration. In salt water, however, it will not stand a week, and 
when exposed to the sun and air, will be destroyed in about a month. 

Each of the fossil resins represents a class that have many varieties, 
but none of them are coniferous. The latter class are those that 
furnish the turpentine and common resins of the present day, which 
are of the least value of any of the resins for a straight varnish or a 
pigment varnish. Their use in a varnish is principally on account 
of their cheapness and the slightly improved brightness they confer. 

Records of the protective nature of some of these varnish paints 
show that a suitable combination of linseed-oil and a resin is a better 
protective vehicle than oil alone, yet the smaller the proportion of 
the common class of resins, the more durable was the coating. 

In the oil and varnish trade, the essential differences in the quality 
of varnishes are due to the kinds of resin used, the proportion and 
quality of oil, and the care exercised in compounding them under the 



* Ureses Dictionary of Arts and Manufactures. 



116 FOSSIL RESINS AND VARNISHES. 

influence of a well-regulated long-heating process. The temperature 
and length of exposure to it necessarily varying with the different 
compositions and quality of the varnish required to meet the condi- 
tions to which it is to be subjected. Heats approximating the charring- 
point of the oil, 450°-500° F., are necessary for a thorough blending. 

Varnishes and varnish paints dry better if moderately warm 
when applied, or if applied to a warm surface. Manufacturers of 
pianos and other highly finished surfaces on wood subject their work 
to 200^ to 250° F. to aid the drying and to harden the coating. 

But cheaper materials and processes than the above are employed 
to produce coatings to compete with the basic metal pigments for use 
on ferric bodies. This careless compounding has resulted in lowering 
both the price and quality of varnish paints, until many of the com- 
mercial varnishes fall below the average of the better class of straight 
pigment oil paints for protective coatings on ferric structures. 

For trade convenience, 100 pounds of resin are taken for the unit 
of composition, and with this unit, 8, 10, 20, or any number of gallons 
of oil rated at 7.8 to 8 pounds per gallon, are compounded for the 
different grades of varnish, known as 8, 10, 12, etc., gallon varnishes. 
To designate the kind of resin used, the initial letter of the kind of 
resin that is employed is taken, viz. : An 8Z varnish means an 8-gallon 
Zanzibar; an 8M, an 8-gallon Manila, and so on, both for the single 
letters or with a combination of the letters. 

. In the color varnishes or so-called enamel or paint varnishes, 
where the pigments are ground in the selected brand of varnish 
employed for the vehicle, the designated letter of the resin in it is 
generally lost or withheld, except as specially furnished by the 
manufacturer. 

All of these varnishes or paints are best thinned with turpentine 
to the proper consistency required for the brush. It is better for 
this purpose than oil. The heating of the oil and resin together for 
the varnish has so thoroughly incorporated them, that no free oil is 
present to exert any change or action in the drying process, separate 
from that present in the coating hs a whole, and which the addition 
of free oil as a thinner would disturb. 

Benzine, or other distilled hydrocarbon liquid, should never be 
used in the composition of varnish or varnish paint. Their quick 
evaporation results in making the coating porous, and liable to 
*' check" or "alligator," as painters term it. 



FOSSIL RESINS, QUALITIES OF. 117 

An essential point in either a straight or a pigment vamish is 
that the linseed-oil should be made from ripe seeds, cold pressed, and 
be well aged, and its "Mucossities" or non-drying elements (nearly 
6 per cent of it) should be removed, in part at least, or so changed 
in character as not to be readily decomposed in the natural oxida- 
tion of the vehicle in the process of drying. 

Dingle^s Jaumal reports the experiments of Dr. Sace (Nurem- 
berg) to ascertain the nature of different resins, viz.: Amber, copal, 
common resin, dammar, elemni, caramba wax, mastic, shellac, and 
sandarach. All of them were reducible to a powder form. Amber, 
elemni, mastic, shellac, and sandarach became pasty before melting, 
the others became liquid at once. Amber and dammar did not dis- 
solve in alcohol. Oopal became pasty, elemni and zaramba wax dis- 
solved with difficulty, while common resin, mastic, shellac, and san- 
darach dissolved easily. Caustic soda dissolved shellac readily 
common resin partially, but had no influence on the other resins. 

Oil of turpentine dissolved neither amber nor shellac; it swelled 
copal, dissolved caramba wax, common resin, dammar, elemni, and 
sandarach easily, and mastic very readily. 

Boiling linseed-oil had no effect on amber, caramba wax, copal, 
elemni, or shellac, while sandarach dissolved slowly; common resin, 
dammar, and mastic dissolved easily in it. 

Petroleum ether had no effect on amber, copal, and shellac, and 
was a poor solvent for caramba wax, common resin, elemni, and 
sandarach, and was a very good solvent for dammar and mastic. 

Benzol dissolved common resin, dammar, and mastic very easily, 
elemni and sandarach to a limited extent, caramba wax more readily 
than elemni, but had no effect upon amber, copal, and shellac. 

Though gums and resins are generally spoken of as belonging to 
the same class, they are distinguished from each other by the solu- 
bility of the gums in water and the insolubility of the resins in the 
same liquid. The gums are insoluble in alcohol, while the resins are 
soluble in it. The so-called gum-resins are soluble in both water 
and alcohol. 

The Trades Journal Review (London), Dec. 4-14, 1901, p. 15, 
announces the discovery by Dr. Kronstein (Hamburg, Germany) of 
the "Synthetical Formation of Vamish Gums and Resins." These 
sjmthetical products are identical in physical and chemical properties 
with those occurring in nature. The discovery includes that of an 



118 FOSSIL RESINS, QUALITIES OF. 

intermediate product between the gums and resins, which invariably 
consists of twelve molecules, affliating with "linoxin," the highest 
oxidation of linseed-oil. Dr. Kronstein produces an artificial resin 
identical with fossil amber, both in color and hardness; also has 
advanced his theory and process by producing the soft resins and 
balsams. 



CHAPTER XI. 

BAKED-JAPAN COATINGS. 

For special locations and ferric constructions, viz. : riveted-steel 
water-pipe lines, anchor- and eye-bars, lattice-trusses, posts and beams, 
covering- or buckle-plates walled in or buried in masonry, and inac- 
cessible for inspection, repairs, or repainting, a special coating called 
*' baked japan" is being tested in a number of locations, the most 
prominent of which is a number of miles of steel water-pipe mains, 
30 to 50 inches in diameter, riveted into a continuous length. 

The process of manufacture and composition of the japan is simi- 
lar to the black-varnish products, but a larger quantity of asphaltum, 
gilsonite, and other cheaper grades of gums and resins replaces the 
finer qualities of fossil and other resins. It is applied by immersing 
a hot pipe or other article in a hot bath of the compound, and upon 
removal from the bath and draining, baking it for a regulated period 
in an oven or muffle kept at an even temperature of 350® to 500® F., 
according to the size of the object to be coated, the composition of 
the japan, and the service required of the coating. 

It fills all small interstices in the object, is elastic, will follow with- 
out strain all changes in temperature of the body coated, is perfectly 
impervious to atmospheric influences, running water, brine, acid, 
and alkaline and sulphur solutions, that affect the ordinary oil-paint 
coatings. Its cost per square yard of coated surface is naturally 
much greater than any brush coating, and will vary according to the 
conditions of its application. Its durability or life may be anywhere 
from ten to fifty times that of the ordinary oil-paint coating exposed 
to the same influences. 

The composition of such baked coatings (and there are scores of 
them in practical use) appears to be of less importance than their 
proper proportion, and the care used in their combination, applica- 
tion, and final baking. It is reasonable, however, to expect that a 
compound that requires a high temperature to apply and to harden it 

will be durable, for the change induced by the heat in the final dry- 

119 



120 BAKED^APAN COATINGS. 

ing of the coating is not alone that due to evaporation, but is a thermo- 
chemical one, that may be supposed will strongly resist any influ- 
ences tending to produce any other change than is subordinate to 
the original one that dried the coating. 

Baked-japan coatings, from the nature of their ingredients, are 
electrically passive, except to currents of high potential; hence it 
remains for time to determine whether the stray electric currents, 
now a fruitful source of electrolysis in all ferric bodies that lie in the 
pathway of their return to their place of generation, will not find the 
rows of rivets that unite the several sections of the underground 
water-pipe lines coated with baked japan, the points to concentrate 
the electrolytic energy for a rapid corrosion of the pipe system at 
thousands of points in each mile, instead of a hundred or so, exposed 
in the usual spigot and bell method of joining the pipes. The elec- 
trolytic action at the rivets will be hastened by the difference in poten- 
tial between the rivets and the pipe-metal — both of which are of 
different potential from the japan coating. The brush-paint coating 
applied to the rivet-heads will afford but little if any protection against 
corrosion or electrolytic action, as they will take place underneath 
the coating, and will require but a small development of either before 
they cast off the paint and have an easier field for their progress. 

Another source of corrosion which these joints can resist but a 
short time is the action of the acid elements present in all earths. In 
the case of these water-pipe lines exposed for miles to a great number 
of strong electric currents, the ordinary rate of corrosion from earth 
and water will be intensified, as in the water-tower stand-pipe case 
cited in Chapter XXXIV of this volume. 

While the baked-japan coating of itself leaves but little if any 
room for improvement in the coating of water-mains, it will 
surely be a source of future regret that a better method of joining 
the short sections of pipe into a continuous line was not adopted than 
the riveted joints thus far used. 

If the brush or modified japan coating applied to the pipe circular 
seams is adequate for their permanent protection from corrosion, 
why incur the expense of a baked coating for the body part of the 
pipe? If it is not a permanent protection for them, then, as a chain 
is no stronger than its weakest link, there must be a great number of 
weak links in this method of constructing and protecting wiiter-pipe 
mains that ought not to be repeated. 

The interruption of the water-supply of a large city is too serious a 



BAKED^APAN COATINGS, 121 

matter to allow the question of a few hundred dollars a mile differ- 
ence between a good and a bad plan of joint construction to be a factor 
in determining which to use. That a number of American cities 
have this bad joint is evidenced from trade catalogues and other 
illustrations of this method of constructing large water-supply pipe- 
lines. 

The question has been raised as to whether the baking of the 
coating effects any further chemical union between the oil and the 
other constituents of the dip, other than that developed in the process 
of manufacture? It is probable that it does, as the baking tempera- 
ture is materially higher than that in the process of manufacture. 
The pipe-coating material before baking is readily soluble in turpen- 
tine, but after baking is not softened by prolonged digestion in hot 
turf)entine, and but indifferently in hot naphtha. The preliminary 
heating of the pipe before immersion in the hot-pipe dip assures its 
adhesion and impermeability, as the air and moisture are practically 
excluded and the preliminary bond of the coating to the metal is 
perfect. The evaporation of the volatiles in the japan dip in the 
process of baking is so quickly effected in the earlier stage of baking, 
that the liquid or fused mass of the resins readily replaces them and 
fills the interstices caused by their evaporation, and ensures a smooth 
unbroken surface to the coating altogether different from that of a 
dried paint. 

The so-called japanned or enamelled coatings used on sewing- 
machines and many domestic machines and utensils are generally 
of that composition that will give the best appearance. They are 
not proof against corrosion under many exposures that would be 
resisted by a good varnish coating or an earthenware salt glaze. As 
a rule they chip easily, and corrosion once established in these spots, 
spreads rapidly beneath the enamel and flakes it off. 

A properly made enamel is essentially a glaze, similar in com- 
position and properties to glass, and has all of the advantages and 
disadvantages of that substance. It is melted at a high heat, 1200° 
to 1400° F., and adheres to the siuiace of metals perfectly. Enamels 
generally resist the action of acid solvents, but are brittle and easily 
chipped off. 

" The best baked japans are intermediate between enamels and 
ordinary varnishes, and resist the action of solvents almost as well as 
an enamel, while they surpass the latter in the tenacity of the coating, 
allowing the metal they cover to be bent to a moderate degree without 



122 BAKED-JAPAN COATINGS. 

injury, while their elasticity is generally greater than a hard varnish. 
In hardness, baked japan is intermediate between varnish and glass, 
or harder than gypsum and nearly as hard as marble." * 

Baked black japans are made from linseed-oil and asphalt as a 
base, mixed with more or less copal resins, usually kauri, and are 
thinned with turpentine. Like varnishes, the more linseed-oil they 
contain and the less driers (oxides of lead and manganese) the more 
durable they are; but to get them to bake hard at a comparatively 
low heat, the proportion of oil is frequently decreased as much as 
possible and the amoimt of driers increased, forming an inferior, 
brittle coating easily injured by a slight blow or rough handling. 

Modem baked-japan water-pipe coatings are very similar in char- 
acter and in their application to Dr. Angus Smith's anti-corrosive 
water-pipe coating, that forms the subject of the following chapter. 



* "Paints, Varnishes, and Enamela*' A. H. Sabin, M.S., New York, 1896 



CHAPTER XII. 

DR. ANGUS smith's ANTI-CORROSIVE WATER-PIPE AND OTHER COATINGS. 

This compound was originally applied by Dr. Smith in 1840, and 
patented in England in 1850, and was first used in America in 1858 
upon some pipes imported from Glasgow. Dr. Smith's original for- 
mula is not definitely known. Mr. James P. Kirkwood's Report on 
the Brooklyn Water Works, published in 1858, gave the following 
formula for it, and it was used to some extent upon the pipes for 
those works; evidently satisfactorily, for Mr. Peter Milne, engineer 
in charge of the extension of the works, reports: "That 36-inch pipe- 
mains laid for 35 years were found to be in perfect condition exter- 
nally, and but few tubercules or other deposits were foimd on the 
inside of pipes." The pipes had been coated by heating them in an 
open furnace to about 500^ F., and then immersing them in a bath 
formed from coal-tar, as follows: * 

Coal-tar was distilled until the naphtha was removed and the 
material deodorized and of the consistency of melted wax or a thick 
molasses. This process also eliminated most of the tarry acids, and 
necessarily required considerable time and care to effect. Five to 
6 per cent and in some cases 8 per cent of pure raw linseed-oil 
was then added and stirred in well. The bath was made deep enough 
to receive the pipes when placed in it vertically. The pipes remained 
in the bath until they had cooled down to the same temperature, 
about 300** F., or about 30 minutes for a 20-inch-diameter pipe. 
Careful attention was given to the length of time the pipes were 
to remain in the bath. A less time than 30 minutes for a 20-inch 
pipe gave an unsatisfactory result. For pipes from 4 to 12 inches 
in diameter, 15 to 20 minutes' immersion appeared to be sufficient 
to get a reliable coating. 

When the coal-tar was distilled to the consistency of mineral 

* " Report in relation to Proposals made by various parties to protect the 
cast-iron water-pipes of the City of Brooklyn from corrosion." By James P. 
Kirkwood, Chief Engineer. City Document, published by Hosford & Co., 1858. 

123 



124 DR. SMITH'S ANTI-CORROSIVE WATER-PIPE CO AT WO, 

pitch or bitumen, or when common resin or Burgundy pitch was 
mixed with it and used as a bath, the pipe coatings became hard and 
brittle when cold, and the bath material would not answer, even 
where the quantity of linseed-oil used in it was increased to 15 or 
more per cent. 

The preliminary heating of the pipes to 500° F. before immersion 
in the bath, after a short experience, was found to be prejudicial, 
and was abandoned. The combustion gases of the heating-furnace 
that were deposited on the pipes appeared to affect the bonding 
of the coating to the pipe-metal, and the pipes when removed from 
the bath were not satisfactory, and new specifications for coating 
them were adopted. 

These specifications required the same preparation of the coal-tar 
for the bath as given above, and for it to be kept at a temperature of 
300° F. during the period of dipping. As the material was continu- 
ally deteriorating during the dipping process, fresh material was to 
be added frequently, and at least 8 per cent of linseed-oil, as near as 
could be guessed at, kept in the bath, or added with the fresh pitch. 
The bath was required to be occasionally entirely emptied of its con- 
tents and to be refilled with new material. The old material after 
a few days' use was found to be hard and brittle like common pitch. 

Every pipe was immersed cold, but not frosty, and was to remain 
in the bath until it had attained the temperature of the bath, 300° F. 
This period was about 30 minutes for the 20-inch pipe, as in the pre- 
vious specification. It required a brisk fire to be maintained imder the 
bath to overcome the cooling action of the cold pipe when immersed. 

The presence on the pipe of moulding-sand, dirt, moisture, frost, 
or oil and grease of any kind, was found to be detrimental to the appli- 
cation of the coating, and their removal was necessary before dipping. 

The royalty paid Dr. Smith for the use of his formula, although no 
United States patent was in effect, was 37^ cents per ton of pipe. 

The price paid the English pipe-founders for coating the pipes 
ordered from them by the Brooklyn Water Works was $1.25 per 
ton, for the years 1858 to 1860. American pipe-founders' and con- 
tractors' price for Dr. Smith's coating was about S3.00 per ton as 
against a plain asphalt coating of $1.83 to $2.25 per ton. 

The efforts of other water-works' engineers to follow Dr. Smith's 
formula, and possibly to improve or cheapen its application, resulted 
in so great a variety of recipes and consequently inferior results, that 
but little dependence can be placed upon their reports. The tem- 



DR. SMITH'S ANTI-CORROSIVE WATER-PIPE COATING. 125 

perature of the preliminary heating of the pipe before immersion 
varied from 200® to 700® F., and the proportion of ingredients and 
their composition was equally startling, as were also the attending 
results. 

Mr. Chas. Harmony, Chief Engineer of the Louisville, Ky., Water 
Works, who experimented for a number of years with Dr. Smith's 
formula as given by Mr. Kirkwood, reports: That ^*some of the pipes 
so coated, after an exposure of from six to eighteen years, were in 
as perfect condition as when first laid; but it was an exception, not 
a rule. In a majority of cases the coating on the inside of the pipe 
was all gone, and upon the outside surfaces it had apparently been 
of no importance in prolonging the life of the pipe. The difficulty 
experienced was, that in the heating of the bath to the temperature 
of 300® F., the coal-tar, resin, and pitch compounds became unman- 
ageable by approximating the condition of boiling and volatilization, 
and going everywhere except in the place it was wanted. The coat- 
ing was thick and apparently unbroken, but exceedingly brittle, 
and would crack and scale off in the ordinary.process of handling." 

The tension of coal-tar and pitch at a temperature of 300® F. is 
hardly less than that of water at the same heat, or equal to about 53 
pounds' pressure. To maintain such a temperature in the bath in 
open atmospheric pressure is impractical, and the composition becomes 
unmanageable. 

Other engineers report that the pipes after twenty years of expo- 
sure were found to be free from corrosion, but the coating had lost its 
bond to the pipe, and evidently remained in place because corrosion or 
other causes had not developed enough energy to cast it off against 
the pressure of the surrounding earth. 

In these and similar instances of failure, the results appear to 
have been more markedly against pipes cast in greensand instead 
of a dry sand or loam-mould, evidently because the thick, vitreous, 
or partly fused greensand coating carried so much air in its rough, 
sandy surface into the bath, that it could not escape through the 
heavy pitch composition. Furthermore, this varnish itself was over- 
charged with its own vapors under tension, and of greater density 
than the air, and confined them until the cooling of the pipe when 
removed from the bath rendered their escape impossible, hence the 
irregularity in the results. Had the pipes been baked in an oven 
after removal from the bath, as in some recent applications of this 
compoimd, the rough, vitreous sand coating on the pipes doubtless 



126 COAL-TAR COATINGS. 

would have ensured a more enduring coating than the same com- 
pound applied to a smooth, dry-sand moulded surface, or one of rolled 
wrought iron or steel. This silicate coating would be reinforced by 
the tough Bower-Barff skin, to which it is naturally so closely attached 
as to require pickling to remove. It is the subsequent baking that 
the pipe receives that renders this process a success. The composi- 
tion of the bath can be varied greatly without much detriment to 
the protective nature of the coating, if the baking process follows 
the bath. 

The generally imfavorable results attendant on the use of Dr. 
Smith's formula without the baking process, and the care and cost 
of it, determined the present practice of the pipe-founders, which is 
to place the pipes for a short time in an oven heated to 250® to 300° F., 
then immerse them in the bath of hot coal-tar and pitch, and then 
cool them in the open air. 

This coating is one of appearance more than of a protective or an 
enduring nature, and is only applicable to water-pipes, as in gas-pipes 
so treated the solvent action of the hydrocarbon vapor soon removes 
the coating, and the joints draw and leak worse than with the uncoated 
surfaces. 

The careless and indifferent boiling of coal-tar, to free it from 
its many acid and other impurities, makes it a variable and unsatis- 
factory coating. Lime, gypsum, and other mineral substances mixfed 
and boiled with the coal-tar to neutralize the ammonia, acids, sulphur, 
etc., only render the tar more unreliable and unmanageable. The 
careless heating of the pipes and bath, also the length of time the 
pipes are left in the bath, and the subsequent treatment of the pipes 
when removed, are all factors in the indifferent results obtained. 

Unless great care is exercised the small pipes will be overheated 
and unequally coated and brushed off, inside and outside. The larger 
pipes, requiring a longer time to heat, from the mass of metal they 
contain, will be underheated in the oven and cool down the bath 
to a lower degree than is requisite for a reliable coating. The subse- 
quent brushing of the coating, both inside and outside, during the 
first period of cooling (a matter of from 30 minutes to 2 hours), 
promotes its reliability. 

All coal-tars or their compounds of whatever nature used as a 
bath, or applied with a brush to any surface, hot or cold, are subject 
to the law of fractional distillation; that is, that such a mixture during 
the process of distillation remains at the boiling-point of that constitu- 



DEAD OIL IN PIPE COATINGS, 127 

ent which boils at the lowest temperature until that constituent is 
exhausted, then changes to the next boiling-point, and remains there 
for a time, and so on. 

The low boiling- or evaporating-point of the lighter elements of 
coal-tar or petroleum products makes them very imcertain in their 
composition, as changes of temperature in the bath from 220^ to 
350° F. are frequently noted without any change in the character of 
them that the eye can detect. 

The character of the bath composition changes so continuously 
and rapidly during the dip that frequent additions of fresh stock 
must be madie. These necessarily cool the bath, change its composi- 
tion, and irregular coatings ensue to that extent that an entirely 
new bath is necessary. 

The use of linseed-oil with coal-tar for pipe coatings, as usually 
applied at the pipe-foundries, is of very uncertain value. It causes 
the dip compound to froth to nearly double its volume, and renders 
the coating lumpy in appearance and uncertain in its bond to the 
pipe-metal. It requires some effort by continual stirring to incor- 
porate it with the coal-tar and pitch, and it is always liable to separate 
from them and float upon the surface, froth, soften the coating, and 
delay its drying. 

Dr. Angus Smith evidently used a number of formulae. for pipe 
coatings that contained linseed-oil as one of the ingredients. A long 
line of careful experiments with the best of coal-tar, pitch, and linseed- 
oil carefully heated and applied, gave almost uniformly good coatings. 
Using the commercial grades of these substances and having the 
ordinary day laborer to compound and apply them, the result was 
necessarily inferior, so much so as to cause the abandonment of 
linseed-oil in coal-tar pipe coatings by modem founders. If, how- 
ever, the truth were acknowledged, the present coal-tar pipe coating 
would be found to be living on the well-earned and deserved reputa- 
tion of Dr. Smith's compound. 

Dead Oil in Pipe Coatings, 

That part of coal-tar obtained in the fractional distillation of 
the tar between the temperatures of 410° to 750° F., and which con- 
tains creosote and anthracine oils (see Analysis, Chapter IX), is 
used to keep the pipe dip at a standard quality. It evaporates ' by 
itself, about one-seventh as rapidly as water, when both are at atmos- 
pheric temperature. 



128 DEAD OIL IN PIPE COATINGS. 

One part of dead oil to about seven parts of coal-tar increases 
the proportion of the heavy oils in the tar dip from about 25 to 35 
per cent, and appears to make the coatings more uniform and of a 
better character than where fresh tar is used to reinforce the bath. 

Thick tar gives thicker and more uniform coatings than thin tar, 
and fresh tar requires a hotter pipe to take bond than does old tar. 

Crude gas coal-tar boiled from five to six hours becomes a soft 
solid at atmospheric temperatures. During the boiling the tempera- 
ture remains at about 220^ F. for about an hour, then rises to about 
290®, stays there for a time, and finally rises to about 350° F. All of 
the naphtha is removed and the tar is deodorized and reduced to the 
consistency of very thick molasses. If to sixteen parts of this tar 
1 per cent of boiled linseed-oil be added, no frothing occurs even at 
400° F. The mixture is thick and does not harden well on light 
iron pipes about }-inch thick. On heavy iron pipes an inch or more 
thick, the coating hardens without diflBculty; in some cases becomes 
too hard, is brittle, and flakes off readily by mechanical injury when 
handled. Dead oD added to thin the mixture causes no frothing. 
The experiments show that linseed-oil could be used with success 
and advantage with partially refined coal-gas tar, and also indicates 
that its application requires more intelligent care than the methods 
employed with the usual crude tar coating. 

Experiments with a refined tar containing dead oil show that as 
high as 8 per cent of boiled linseed-oil resulted favorably in solidity 
and hardness of the coating. In other instances, where from 1 to 
8 per cent of raw linseed-oil was used instead of boiled oil, frothing 
occurred and a poor coating resulted, evidently due to the presence 
and evaporation of the water in the raw oil. There is about 5 
per cent of water naturally held in combination with the best quality 
of raw linseed-oil made from ripe flaxseed, and nearly 8 per cent 
in the oil made from \mripe seed. With many brands of commercial 
linseed-oil, 10 per cent additional of water is frequently incorporated 
by stirring it in with a paddle or passing it through a mixing mill. All 
such oils are likely to be made up from fish, resin, mineral or vegetable 
and animal oils with no linseed-oil of any quality in them, and all the 
difficulties developed by these mixtures with the coal-tar are saddled 
upon the scapegoat, linseed-oil. 

Refined coal-gas tar is practically out of the market and has been 
for many years. What small amount of crude coal-tar is available is 
too valuable and in too great demand for the chemical products in it 



WATER-PIPE DIPS AND COATINGS. 129 

to allow of its use to the great extent that pipe-founders require for 
their work. Heavy roofing pitch alone will run in moderately warm 
weather and becomes too soft and sticky for a pipe covering, unless 
laid immediately after coating. This is impracticable, and in cold 
weather it is too hard and brittle for transportation or handling. 

Nine parts of heavy roofing pitch with one part of boiled linseed- 
oil give a thick glossy coating less brittle than pitch alone. 

Two parts of boiled linseed-oil with the nine parts of the pitch 
give a coating more elastic and tough. 

Three parts of boiled linseed-oil with nine parts of pitch, the coat- 
ing is more bulky and less smooth than with the others, while with 
larger proportions of the linseed-oil the coating partakes of the char- 
acter of a slow-drying paint and requires baking, which gives it a 
superior quality. 

Coal-gas tar belonging to the class of pyrogenic (fire-formed) 
compounds is unstable at ordinary temperatures, and is continuously 
decomposing by the evaporation of its many hydrocarbon elements, 
until nothing but the hard friable pitch is left, which contains nearly 
all of the sulphur element in the coal that forms the base of the tar 
product. Asphaltum, also a pyrogenic product, formed by the slow 
evaporation or distillation of petroleum, decomposes upon exposure 
by reason of the oxidation of the sulphur element in it, but is more 
durable than the coal-tar residuum or pitch (asphalt). 

Asphaltum and linseed-oil coatings do not harden well, unless a 
hard grade of asphaltum is used. 

Water-pipe Dips and Coaiings. 

There are many pipe dips upon the market, some covered by pat- 
ents of doubtful validity, others secret or proprietary compounds of 
doubtful utility. Some of these compounds appear as pipe dips, 
also as brush paints appUcable for ferric constructions other than 
pipes. (See Paint Tests, Chapter XXIX.) 

The P. and B, Pipe Dip is a patent dip; the principal ingredients 
are probably an asphalt and candle-tar pitch. The latter is a pitch 
obtained by the distillation of animal fats or refuse. Upon pipes its 
coating is similar to an asphalt coating, not hard nor brittle, not 
very glossy nor very tenacious. 

The P. and B, Universal Paint is an asphaltum paint; the 
vehicle contains carbon disulphide as the volatile element, the evapo- 
ration of which is not only nauseating and dangerous to all animal 



130 water-pipe dips and C0ATIN08. 

life, but carries the danger of explosion and fire risk. Whatever good 
qualities it may possess when on, are more than offset by the dangers 
connected with its application. 

The P. and B, ''Rvberine^' consists of "ruberoid" dissolved in 
naphtha. "Ruberoid" consists of California asphaltum or maltha 
and candle-tar pitch digested and vulcanized with sulphur. "Ru- 
berine" dries rapidly, is hard to spread smoothly, but gives an elastic 
or rubbery coating. See tests of paints, New York Elevated Railway 
Viaduct, for an example of its qualities. 

Mineral RvJbber Dip (or Rubber Coating) is a secret composition 
whose appearance indicates that it is largely asphaltum. It is rather 
duller in appearance than the ordinary coal-tar or asphalt mixture. 
The dip requires a temperature of about 400^ F. to apply, and then it 
is almost impossible to get a smooth or neat-appearing surface. As 
yet its protecting qualities have not been determined. 

'^ Biturruistic'' Products comprise an enamel to be applied in a 
molten state to the metal. Bitumastic cement is used hot for the 
preservation of ships' bilges and frames, instead of the usual hydraulic 
cement coatings, and also for the protection of water-pipes. Bitu- 
mastic solution has bitumen for its base. It is a brilliant black 
paint applied the same as other paints, and is probably similar in 
character and composition to "Smith's Durable Coating." It has 
been used to a considerable extent on steel water-pipes and for 
the limited period of test in that service is favorably spoken of. It 
dries in 24 hours, is said to be unaffected by acidulous, alkaline, 
or brine solutions. If applied to the clean dry surface of the metal, 
does not crack or peel when alternately wet or dry, or exposed con- 
tinually to running water in penstocks, water-wheels, etc. It is 
not affected by a moderate heat, nor by sulphur fumes, and is fur- 
nished ready for use at $1.75 per gallon. It is very volatile, and the 
packages must be well stirred while being used. Its covering power 
is about 400 square feet, and its weight about 9.5 pounds per gallon. 

"Crysolite'^ Enamel and Paint, "Crjrsolite" paint is made from 
oil and a by-product, oven-coke. It weighs 9.5 pounds per gallon 
and spreads 500 square feet as furnished for a paint. When thinned 
with 12.5 per cent of oil, will cover 1000 square feet, and under general 
conditions in both hot and cold weather, dries completely in 30 hours. 
**Crysolite" Enamel is a quick-drying paint of the same character as 
the above, and dries in one hour. *'Crysolite" products are alkaline 
in reaction, whereas coal-tar products in general have an acid reac- 



WATER-PIPE DIPS AND COATINGS 131 

tion. **Crysolite" is better for being applied warm or hot (as all 
ferric paints are). In the winter one-eighth of its volume of tur- 
pentine can be added to aid its spreading power, which can be made 
to cover from 800 to 1000 square feet. **Cr}^solite" paints cost 
about 75 cents per gallon mixed ready for siunmer use. 

**Crysolite" coatings on ammonia tank-cars and reservoirs stand 
the action of ammonia liquors and gases better than most of the paints 
used for this purpose. *'CrysoUte" baked coatings under test re- 
sisted the action of carbonate of ammonia and ammonium chloride 
liquors for three months without injury. "Crysolite" under the 
influences of strong brine is more favorable than the commercial 
asphaltum or the ordinary coal-tar paints. 

Hickenlooper'8 gas-pipe-dip compound, used by the Cincinnati 
Gas Light and Coke Company, the United Gas and Improvement 
Company, and several other gas companies, to coat their small service 
pipes, has a record of many years'* exposure in the ground with few 
traces of corrosion. .The failures thus far reported show that neither 
the process nor compound were at fault, but the lack of thoroughness 
and intelligence in its application. The pipes are first cleaned from 
rust and mill-scale and then immersed in the following dip and 
in the following manner. Twenty gallons of retort coal-gas tar are 
brought up to a boiling heat for a short time to evaporate as much 
of the water, acids, ammonia, etc., as possible, then 20 pounds of 
freshly slaked lime are sifted in from the top and well worked into 
the tar. Boil down to the consistency between a coal-tar and a 
pitch. When settled, add four pounds of tallow and one pound of 
powdered resin; stir until all are dissolved and thoroughly incor- 
porated, then let the mass cool and settle; then ladle oflp into barrels. 
When ready for use, to each barrel of forty-five gallons of the above 
mixture add four pounds of crude india-rubber dissolved in turpen- 
tine to the consistency of thick cream. Heat the mixture to about 
150° F. and immerse the pipe, previously heated to about the same 
temperature. After a few minutes' immersion the pipes are removed 
from the bath and laid upon skids to harden. The coating is some- 
w^hat softer than the usual pipe-founders' dip, and requires more 
time to harden, and continues hardening for a number of hours after 
cooling down to atmospheric conditions. The compound is especially 
useful in coating the screwed ends of threaded pipes. It is better for 
this purpose than the red-lead compounds usually employed. 

All rough coatings are detrimental to the life of water-pipes. 



132 WATER-PIPE DIPS AND COATINGS. 

Upon the inside surfaces the pits or cavities that constitute the 
rough surface of the coating are the first to catch the saline, sulphur, 
or other impurities in the water tliat form the basis for the develop- 
ment of the rust cones. The coating under these pits is the first to 
break down, being of inadequate thickness — probably only ^^^ inch 
thick. The external surface of the pipe is as rough as the inside, 
and is not only exposed to the moisture to inaugurate corrosion, 
but this moisture will contain all the acids in the soil in which the 
pipes are laid. 

in all cases of the corrosion of water-pipes, it is the porosity of 
the coating that causes the formation of the tubercles and decay of 
the pipe. Nearly all of the dip coatings, when tested by themselves 
or not in contact with ferric substances, were practicall}^ uninjured 
by acid solutions or running water. 

In general, all pipe coatings, applied as they nearly always are 
in a careless, indifferent manner, 'will begin to show indications of 
tubercles in three years, and cases of tubercles in large pipes at the 
end of sixteen years have been noted, where the carrying capacity 
of the pipes had been reduced 20 per cent.* Engineers must earnestly 
take up this question of reduced carrying capacity of their water- 
pipes and decide whether it is not more economical to add from 5 to 
10 per cent to the cost of the pipe in the form of better coating mate- 
rials and better methods of their application than to submit to this 
decrease in flow, that always grows less with the age of the pipe, wliile 
the demand upon the service is always increasing. 

Specifications for pipe coatings appear to be of little use in produc- 
ing a satisfactory coating, either in appearance or durability, as the 
directions they give are more often evaded than carried out by the 
foundry employes. After the pipes are coated and upon the drying 
skid, no ordinary inspection can determine the character of the coat- 
ing other than its appearance to the eye or touch. 

Testing pipe coatings is usually by the hammer to see whether 
the coating is so hard and brittle as to chip off in handling. The acid 
test determines the porosity of the coating by attacking the metal 
through the pores of the dip. A solution of one part muriatic acid and 
two parts water will affect both the coating and the covered metal 
more at the end of sixty days than they would be affected by ten 
years' exposure to running water. In nearly all cases where the 

♦Excerpts from a paper by Desmond Fitzgerald, C.E. Transactions Amcri- 
oan Society Qvil Engineers, Vol. XXXV, 1896, p. 241. 



WATER-PIPE DIPS AND COATINGS. 133 

coating has been dried by heat or baked, the metal will be corroded 
^ inch or more, the coating undermined and peeled off. 

After all, in this age of specifications, inspections, sciimping, and 
adulterations, there is nothing equal to an honest and capable con- 
tractor, either for furnishing pipe, coating, inspecting, or laying it. 
Get such a one if possible and then watch him closely. 

Generally, the time that the pipes are left in the hot bath does 
not exceed one minute, and is more often only one-half a minute. It 
is impossible to properly coat a pipe in one-half a minute, as the air 
carried into the thick turgid bath by the pipe will not escape in that 
time, and the top part of the inside of the pipe and the lower part of 
the outside of the pipe are uncertainly coated for this reason. The 
pipes are seldom turned over while in the bath, or outside while on 
the skids in the process of scraping and brushing off the surplus dip. 

On pipes that are left in the bath for five minutes the coatings 
are markedly superior to those exposed for shorter periods. This is 
the case whatever the nature of the coating, and is one reason why 
the Angus Smith and other older-day coatings gave such superior 
results to those coated by modem methods. They never had less 
than five minutes in the bath, and were often left for fifteen or even 
more in case of large pipes one inch or more in thickness. Modem 
pipe-foundry management allows no such exposures. 

A coal-gas tar paint that has given very good results in the coating 
of gas-holder tanks and other situations where the metal is exposed to 
ammonia and sulphurous acids in solution and to alternate melting 
and drying under a great range of temperature, is made as follows: 
Coal-gas tar is well boiled to evaporate the water and light hydro- 
carbon elements and then 20 to 25 per cent of caustic quicklime is 
sifted and well stirred in to neutralize the acid elements in the tar. 
This is to be kept hot for a few hours and then an equal quantity 
of good Portland or hydraulic cement is sifted and stirred in thor- 
oughly. The mixture is applied hot to the clean dry iron, and can 
be repeated soon as cool or dry if the exposure conditions are to be 
very severe. In the latter case, a little more cement should be added, 
so that the caustic lime and cement mixture will contain 50 per cent 
of each. The pigments thicken the coal-tar and prevent it from mn- 
ning under sun temperatures and give a bond to the bmsh coating of 
neat Portland cement that should be applied to the coal-tar coat as 
soon as either the first or second coat of the mixture is dry. This 
coating can be repeatedly applied with advantage. It is impervious 



134 PlPE-DtPPING TANK. 

to gases and water and has no tendency to run at temperatures 
under 130° to 140° F. 

Mr. Bom, in " Comptes Rendes," in 1837, called attention to the 
fact that iron cast in charcoal-coated or chilled-iron moulds was less 
susceptible to corrosion than greensand castings. 

The city of Perth, Scotland, where very pure water is obtained 
from the Tay, had their water-pipes coated with a solution of india- 
rubber. After 25 years of use every pipe under 5 inches in diameter 
had been completely closed by corrosion. la 
many cases where the ordinary coal-tar dip 
had been used on the water-pipes it scaled 
off in strips and was discharged at the house 
service-taps. 

A pipe-dipping tank being required for 
some steel riveted pipes, 16 to 30 inches in 
diameter and 28 feet long, was extemporized 
from old material in the contractor's yard, 
and is shown by the following Fig. 21.* 

An old boiler-sheil 3 feet or more in diameter 
and 26 feet long was fitted with a slightly 
dished wrought-iron flange 2 feet or more in 
width all around, riveted to the top end of 
the shell. This served as a working platform, 
also to catch and return the drip. The other 
end or lower one of the shell was riveted and 
caulked to a casHron plate-head which car- 
ried on its inside face a concentric flange in 
the centre, to which was riveted steam-tight 
a wrought-iron pipe nearly as long as the 
outer boiler-shell. This inside pipe was clased 
steam-tight by a conical head that also served 
as guide for the pipe when it enteretl the bath 
of pipe dip. The bottom flange was tapped 
for steam- and drain-pipe connections, which 
were fitted with the usual gates, worked from 
Fio. 21.— A pipe-dippins the surface of the ground, in or on which the 
tank. tank was erected. The annular space be- 

tween the centre pipe and shell was fille<l with the coal-tar or other 
pipe dip to be applied to the pipe, which was kept hot by the steam 
• Engineering Record, Vol. XXXV, May 8, 1S97, p. 489. 



WATER-PIPE DIPS AND COATINGS. 



135 



in the centre pipe. The immersion and withdrawal of the pipe to be 
coated kept the bath mixture well stirred up and ensured a nearly 
uniform quality of its ingredients. It is obvious that this compara- 
tively inexpensive device is adaptable for many occasions that would 
not warrant a more expensive plant. 

A larger shell could be fitted with a number of the conical- 
headed pipes with their separate steam- and drain-pipe connections 
and be available for dipping a number of pipes at the same time, and 
would certainly ensure a more reliable coating than where the pipes 
are immersed in a long horizontal tank. 

Approximate Relative Cost op Various Pipe-Dips and Coatings.* 



Coating. 



Crude tar 

Pitch 

Pitch and linseed-oU. . . 

P. & B. dip 

Mineral dip 

P. & B. universal paint. 

P. <fe B. ruberine 

Tar varnish 

Dutch varnish 

Sabin's baked japan . . . 



Approximate 
Amount Re- 
quired to 
Coat One 4&- 
inch Pipe, 12 
Ft. Long. 



3} gals. 

c tt 



20lbs. 
20 " 
1) gals. 

14 
11 
IJ 
14 



(( 
tt 



Approximate Pricea. 



$3.00 per bbl. (62 gals.) 
5.00 



tt 



tt 



tt 



ft 



45.00 per ton 
75.00 " " 
1.00 per gal. 
1.00 " " 
0.10 " " 
0.25 " " 
1.75 " " 



Cost of 
Material for 
One 4S-inch 

Pipe. Ap- 
proximately 

325 Sq. Ft. 
of Surface. 



SO. 22a 
0.50 

.70 

.45 

.75 
1.506 
1.506 

.156 

.40 6 
2.60c 



a. About 30 per cent of this was lost by evaporation. 

6. Estimated cost of this coating as applied with a brush. The wastage 
would be excessive as a dip, but the dip is the only practical way for its use on a 
large scale, hence the figures are not strictly correct. 

c. This coating requires a comparatively expensive plant and considerable 
skilled labor, which would largely increase the total cost. 



* " The Manufacture and Inspection of Cast-iron Pipes." Thos. H. Wiggins, 
C.E., Boston. Civil Engineers* Association Journal, 1899. 



CHAPTER XIII. 

GRAPHITE AND GRAPHITK PAINTS. 

Carbon assumes in nature three allotropic forms, viz. : Diamond, 
graphite, and amorphous carbon. Graphite itself assumes different 
forms, some of which are amorphous and others strictly ciystalline in 
character. 

If the three allotropic forms of carbon had each a characteristic 
name, no confusion would be liable to arise in speaking of them. We 
speak of the diamond and of graphite, and each is clearly defined. 

In speaking of the third form we are limited to amorphous carbon. 
This form is found in certain stages which are not strictly amorphous 
or granular in character. Coke, for instance, is one form; the others 
are the mineral graphite-carbon or graphite, termed foliated (flake), 
amorphous (granular), etc. Graphite is found in many parts of the 
world and Is of various degrees of purity, ranging from 60 to over 90 
per cent of graphitic carbon in the foliated form and 20 to 60 per cent 
in the other forms. 

The foliated is a designation for the thicker flakes in the Ceylon 
and like varieties, while flake is used to designate the thin flakes of 
the purest brands, similar to the Ticonderoga mine product. 

The German (Bavarian), Siberian, Mexican, and some American 
varieties are amorphous and vary greatly in the amount of carbon in 
their composition, as will be seen from the follo\^ing analyses: 

The purest brands (Ticonderoga mine) have a specific gravity of 
1.21 to 1.4. The amorphous varieties range from 1.80 to 2.25 to 2.79. 
When pure it is perfectly opaque, iron-black or steel-gray in color, 
with a metallic lustre. Its hardness varies from 1 to 2, and it con- 
ducts electricity nearly as well as the metals. 

Pure graphite or minerals high in graphite-carbon grind and feel 

greasy, and are repellent to moisture and oil. Flake-graphite above 

80 per cent in purity, by long trituration with water, can be reduced 

to a fine lamina or pigment. 

Anthracite coal is an intermediary carbon between graphite and 

136 



GRAPHITE AND GRAPHITIC CARBON, 137 

bituminous coal. It is blacker than graphite, hardness 2 to 2.1 as 
against 1 to 2 for graphite when it contains 95 to 99 per cent of carbon. 

The Ceylon, Cumberland, Indian, and American flake varieties 
are the purest in carbon, and are used for pencils, crucibles, lubricants, 
stove-polish, foundry facings, etc., and to tone up the poorer varieties 
for many purposes. 

Foliated graphites, though used for pigments, are not as satisfac- 
tory (for reasons given hereafter) as the amorphous variety, that, less 
rich in carbon, contains other mineral substances, non-corrosive, non- 
absorbent of atmospheric moisture and gases, either as individual sub- 
stances or collectively as a natural mineral compound. That this 
feature may be duly considered when a graphite pigment is to be 
selected for a ferric structure the following analyses of amorphous 
graphite from a number of widely separated mines are given: 

Analyses of Amorphous Graphite. 

Siberian and German Mines. \f^ A^MinS 

Per Cent. Per Cent. 

Graphitic carbon 33.20 to 36.06 28.39 to 33.48 

SUicaasSiO^ 43.20 " 37.70 46.97 " 37.54 

Iron soluble as Fe.0, i ^ ^ ,, ^^ ^ 22 « 4.25 

'' insoluble ) 

Alumina as AljO^ 15.42 " 17.80 16.90" 12.35 

Calcium as CaO 

Magnesia as MgO 

Carbon dioxide, water combinect 
dium compounds, iron pyrites, voLa- 
tUe matter and Joss 4.00'' 8.22 2.53" 1.36 

Specific gravities, 2.25 to 2.79. Color, gray or drab. Hardness^ 
1.5 to 1.8. Fracture, granular. 

Graphite from the Wisconsin mines analyzes, viz. : 

Graphitic carbon 72.00 to 74 .00 per cent 

Ironoxide 7.10 " 14.00 " " 

Silica 10.00 " 12.00 " " 

Alumina 8.00 ' traces 

Water and undetermined 2.90 " " 

The Mexican graphites are amorphous in character, are high in 
carbon, and have had but a limited use for pigments. When contain- 
ing about 80 per cent of carbon they are better suited for lubricants 
or foundry facings. 

The amorphous brands of graphite require no calcination other 
than necessary to dispel the ivater, natural to all minerals, prepara- 



I 1.06" 1.20 0.99 " 1.02 



138 GRAPHITE AND GRAPHITIC CARBON. 

tory to any pulverizing operations. They grind fine and granular, 
and in an approximately cubical form, are not repellent to the oil 
or vehicle, and are nearly as unoxidizable from moisture, atmospheric 
influences, combustion, and other gases as pure carbon. They are 
of an agreeable color and good covering power, and they work well in 
combmation with other pigments; flow, hold, or carry well in the oil, 
and are as easily brushed out to cover as much surface as any good 
paint. They are not repellent to the oil, do not separate from it, 
nor set in the paint pot or barrel on long storage, either as a paste or 
paint. They are wholly self-supporting as pigments, contain no 
elemental substances that tend to reduce them to a lower plane by 
oxidation or slacking in the presence of moisture and gases. They 
require no body stuffing, either to bond them, or to keep them quiet, 
or from curdling or crawling during or after application, and they 
contain neither acids nor sulphur. 

They are entirely diflFerent in character and composition from 
the so-called silica graphites of commerce, many of which resemble 
carbonaceous schists or impure soapstone, or are compounded from 
flake-graphite and mineral substances of dissimilar character, such 
as barytes, silica, furnace slag, etc. These several substances, even 
if they are non-corrosive, or electrically or chemically passive of them- 
selves or collectively, when assembled in a paint cannot be as reliable 
as are the same substances incorporated together by the processes 
of nature, each and every particle of which is of the same physical 
and chemical composition and equally affected by the vehicle, 
atmosphere, or other conditions that affect a paint. 

They have not the merit of being synthetical compounds. No 
human care in the mechanical processes of grinding and mixing them, 
as a compound pigment or paint, can arrange them in sequence or 
in other than a haphazard manner. 

Silica graphite paint is of a dark, lifeless brown ; not objectionable 
on the enclosed ironwork of a building, but decidedly so for more 
prominent positions. Hence it is toned up by red lead or other basic 
pigments of agreeable color, but at an increased cost and a contributed 
element of danger in the disintegrating of the paint whenever hydric- 
sulphide fumes reach the red lead in the coating. 

Iron oxide is also used for toning effects, but the natural red hema- 
tite oxides are not strong enough in color to materially modify the 
dark brown of the graphite, silica, and barytes compound, unless excee- 
sive quantities are used, which bring into the coating all of the uncer- 



GRAPHITE AND GRAPHITE PAINTS, 139 

tain elements of that class of pigments that thus far have proven to 
be the most unreliable of all ferric coatings. The danger is greater 
if the brighter red copperas oxides are used. Their strong sulphur 
element sets into action an antagonism between every element in the 
coating and delays the drying of the paint, making necessary exces- 
sive amounts of strong driers to counteract even a small percentage of it. 

Graphite paints are noted for being slow driers and require a lib- 
eral use of driers to get a firm coating. This is more apparent with 
flake-graphite; its flocculent form and oily nature prevent the vehicle 
from bonding it. There is a movement in the paint during the whole 
process and period of drying that even the sharper and more angular 
form of the silica or barytes added cannot wholly overcome. Further- 
more, these substances bring their own peculiarities into the coating 
and forcibly demonstrate the unreliable character of all compound 
paints. The greater the number of substances in a paint the less 
dependence can be placed upon them to work together for a durable 
coating. An acid and an alkali will chemically form an innoxious 
whole, but this or similar action is dangerous in a drying paint and 
generally proves detrimental to the coating or covered surface. 

High-carbon graphite is so easily adulterated with soapstone that 
if a pound of it be ground with three pounds of soapstone (specific 
gravity 2.7), neither the eye nor touch can detect the adulteration; 
only analysis will show it. 

Graphite is one of the lightest pigments. Its specific gravity 
ranges from 1.21 to 1.4 to 2.38, while zinc oxide is 5.42, asphaltum 
1.4 to 1.8, barytes 4.5 to 4.7, silica 1.9 to 2.8, gypsum 2.15 to 2.35, iron 
oxide 4.7 to 5.4, whiting 2.2 to 2.8, red lead 9.07, white lead 6.43. 

The natural drying of a linseed or varnish coating is in the form 
of a closely woven web of a fine fabric. This shows plainly on a freshly 
dried or drying surface, and explains the reason why two or more 
coats are necessary to give a smooth foundation for the last or polish- 
ing coat. Each subsequent coat fills the interstices of that under- 
neath it, each coat repairing the other's deficiencies, as many folds of 
a fine muslin will in the aggregate make an adequate covering from 
heat or light. 

Now, it is the function of a pigment to fill these cellular forma- 
tions in the drying vehicle, or rather, while being applied with a brush, 
for the atoms of the pigment, mechanically arranged in brushing out 
the paint, to lie side by side, all embedded in the vehicle, which in 
drying naturally takes the lines of least resistance, i.e., between the 



140 DRYING OF GRAPHITE PAINTS 

pigment atoms, and, as it were, each atom lies in an approximately 
square hole, the most favorable condition for the bond between the 
pigment and the vehicle. 

If the pigment atom be splintered like a sliver of glass, or of only 
length and breadth like a flake, then the natural cellular formations 
in the drying vehicle cannot be realized. Such shaped pigments are 
arranged with the sharpest angles and edges upright to the drying 
surface, and are not well covered in or embedded in the oil, hence 
dry with a rough surface that will hold moisture and dust and quickly 
decompose and disintegrate them from their bed, when more moisture, 
cinders, and dust take their place and the cycle of action is repeated. 

The rough character of all paint coatings containing siUca, barytes, 
furnace slag, etc., is distinctly apparent to the touch. A round marble 
does not bed itself in a cement as well as the cubical block from which 
it is made, neither does a beach-worn sand or a quicksand atom, 
with the best of cement, make a good mortar for the same reason. 
The splinters, flakes, and roimd atoms are more easily removed from 
their beds than a square atom. 

Both the amorphous and flake graphite pigments (not associated 
with foreign substances as adulterants) being electro-negative, are 
less affected by catalytic or electrolytic action caused by the juxta- 
position of electro-positive substances in the coating or surface 
covered, or by hydro-sulphide gases, than any class of pigments, 
lampblack alone excepted. This is a valuable feature in any paint, 
whether applied to iron or wooden bodies, and in the future will insure 
a more extended use of graphite paints instead of the iron oxides, 
and compounded or patent paints, to the careless use of which most 
of the corrosion in progress upon important ferric structures is di- 
rectly traceable. 

As a general rule graphite minerals that contain about 40 per 
cent of graphitic carbon have proven to be better for pigments than 
those .richer in carbon, for the reasons given before, the principal 
one being that they are less repellent to the oil and bond better to it, 
and do not appear to be affected by combustion gases. 

Amorphous graphite coating applied to boiler tubes exposed to 
internal firing and the action of hot water under pressure of eighty 
or more poimds per square inch for two years was uninjured and 
fresh as when it was first applied. 

Fig. 22 shows a boiler-tube thus exposed. The tubes had been 
in the boiler for a number of years, when they were removed and 



DRYING OF GRAPHITE PAINTS. 141 

cleaned from the hard silicious scale that covered them. When they 
were replaced, each alternate tube was coated with Lake Superior 
graphite paint, the others being uncoated. At the end of over two 
years a number of the tubes were removed and their condition as- 
certained. The unpainted ones were again covered with the hard, 
flmty scale, that required the use of a boiler-scraper to remove,' The 
painted tubes were covered with a light flocculent coating of the scale, 
that could be brushed off with the fingers, showing the bright, clean 



Fig. 22— Boiler- tube. 

paint beneath it. The tubes were pitted with rust in spots and 
streaks when they were first removed to be cleaned. These show 
in the photograph, b\it the corrosion was stopped by the paint. The 
light-colored scale-deposit was left on part of the tube, and shows 
on the sitles of the figure. 

Pieces of iron coated with graphite paint and dipped in muriatic, 
sulphuric, and nitric aci{ls, and allowed to dry with the acid on them 
showed at the end of nineteen days no injury to the coatings. Other 
samples immersed in alkaline and other chemical solutions, also in 
coal-oil for a number of weeks, and strong brine for months, showed but 
little injury. Other tests of endurance of all brands of graphite paints 



142 ELECTRIC-FURNACE GRAPHITE. 

showed a marked superiority over other basic pigments, whether 
prepared for a test or compared with the commercial brands of other 
paints. 

While tests of paints are not regarded by many engineers as 
indicative of their value to resist the ordinary influences upon a coat- 
ing exposed to weather, they do show that a coating that can withstand 
the above severe tests is certain to give more satisfactory results 
in its general use than the many commercial paints whose low price 
and not their protective qualities is their principal recommendation. 
They also show that if the conditions to which a coating is to be 
subjected are kno\^Ti, it can generally be furnished to successfully 
meet them. 

Other tests of graphite paints in competition with other commer- 
cial paints are given in the following chapters on paint tests. For 
a roofing paint the graphites high in carbon, of themselves, or mixed 
as silica-graphite compound paints, are of marked excellence. They 
do not harden as rigidly as the iron oxides used for roofing purposes. 
They endure long exposure to the sun, hence are less liable to crack 
or flake off, and they follow without injury the expansion changes 
in the metal they cover. Their darker color and higher cost com- 
pared with iron-oxide paints used for roof coatings are more than offset 
by their better protective qualities and longer life. 

ElectriC'fumace Graphite. 

However engineers may differ about the respective merits of a 
low or medium grade of amorphous-graphite mineral for a straight 
paint compared with a flake-graphite and silica compound, their 
attention is liable to be attracted in the future to a new product 
that has entered the field for a pigment, under the name of 

^^Acheson Graphite.** 

This substance is an amorphous-graphite pigment of high-carbon 
content, whose physical character seems to be materially different from 
the high-carbon mineral graphites heretofore used for paints. Although 
amorphous or granular in character, as compared with the other 
forms of graphite, such as the flake-graphites, it is nevertheless dis- 
tinctly a graphite product, and contains absolutely no trace of amor- 
phous carbon, the name usually applied to such forms of carbon as 
lampblack, coke, coal, etc. Graphite in any form is much more inert 
chemically than these amorphous carbons, but Acheson graphite is 



ACHESON GRAPHITE. 143 

even more inert than any of the natural graphites. Its specific grav- 
ity is 2.25. It is ground dry and air-floated to an exceedingly fine 
state of subdivision and of great uniformity in size of the individual 
particles. 

Its amorphous character renders it far less repellent to the oil 
than the natural graphites containing approximately the same per- 
centage of carbon. This quality causes it to remain in place in the 
oil, and it is not as easily moved out of position by the drying action 
of the vehicle, as is the case with a high-carbon flake-graphite. 

Acheson graphite used with a boiled-oil vehicle will set in the 
coating without the aid of any inert substance to hold it in place 
while drying. Used with raw oil, it requires a drier to secure the 
initial set of the paint, particularly if the coating is to be an ex- 
ternal one exposed to the vicissitudes of weather. 

Its manufacture is entirely unlike that of any other pigment, and 
is shown by Fig. 23, illustrating the style of the special electric furnace 
used to produce it. 

In manufacturing graphite in this way, anthracite coal is heated 
several hours in the electric furnace by means of a powerful electric 
current, approximating 1000 horse-power of energy. The tempera- 
ture of the mass of coal is raised to a point where the carbon is con- 
verted into carbides of the various constituents of the ash, which 
in anthracite coal are very evenly distributed. The temperature is 
then carried to the point where the carbides are decomposed, and 
the principal constituents of the original ash, silicon, iron, sulphur, 
aluminum, etc., are driven oflf as vapors. 

The residue removed from the furnace is carbon in the form 
of graphite, perfectly free from any trace of the amoiphous carbon 
or coal from which it was produced. Its method of manufacture is 
probably a duplication, upon a small scale, of the process by which 
the natural graphites were formed in the earth. The purity of the 
product depends upon the temperature to which it has been raised; 
for commercial purposes, it contains about 90 per cent of carbon. The 
10 per cent of ash still remaining in the carbon is practically as inert 
as the graphite itself, and intimately associated with it. The furnace 
product is broken up, and the grades suitable for various purposes 
separated. The grade used for a pigment is pulverized to an im- 
palpable powder, and air-floated for collection. In the latter stage 
it contrasts strongly with the poorly ground and coarse particles of 
many of the mineral-graphite pigments, and like a properly made 



144 AC HE SON GRAPHITE. 

lampblack or a sublimed-lead product, it has the physical character 
for an ideal paint. 

Unfortunately, there is no standard for a graphite paint as there 
is for a pure-white or red-lead paint. The consequence is, that where 
graphite paint is specified by the engineer, he is to an extent working 
in the dark, and does not feel at all sure but that the coating will be 
spread from some one or other of the many abominations under the 
guise of "mixed paint/' that has not an atom of graphite of any 
kind in it. Reputation of the manufacturer or dealer in graphite 
pigments or paints is quite as essential as in the case of the lead and 
zinc products. Adulterations in a graphite-mixed paint are more 
easily concealed from the eye than in those having a lead or zinc 
base, and are equally, if not more, annoying to the engineer. 



CHAPTER XIV. 

BESSEMER PAINT. 

A SPECIAL pigment, claimed to be of the inert class, has lately 
come into use to replace oxide of iron as a straight paint for ferric 
structures. During the short period it has been upon the market, its 
use has been attended with many favorable results. It is a German 
development, and is reported to be the pulverized slag from Besse- 
mer basic process steel furnaces. It is claimed to be free from the 
sulphur and phosphoric acid elements that are usually present in 
iron-oxide pigments. 

It is prepared as a mixed paint ready for use. The finely pulver- 
ized furnace slag is ground in linseed-oil containing a small amount of 
one of the copal resins that makes the paint coating very elastic 
even after long exposure to the sun. 

It is claimed to spread easily, covering 1000 square feet per gallon; 
but to do this the use of short bristle brushes is recommended, the 
efifect of which is to rub out the coating very thin. But however 
closely the paint may thus be forced into contact with the surface 
being covered, it cannot be as well protected as where the paint is 
spread by long bristle brushes, and a sufficient amount of painters' 
labor and time is given to spread the coating. The pigment is not so 
deeply embedded in the vehicle, nor so well protected or bonded to 
the coated surface as it is when spread over a smaller area. 

Bessemer paint in its natural color is a very dark gray, though it 
can be made a lighter shade by the addition of other substances (not 
of its own nature) to tint it. In this case the coating will be no more 
durable than the life of the most perishable pigment in the paint, as 
is the case with all compound paints. 

Bessemer paint weighs 12^ pounds and costs $1.50 per gallon, 
and is claimed to drv in 24 hours without the use of added driers, which 
seems to indicate that the vehicle carries an energetic drier not 
natural to a linseed-oil and copal vehicle, as the pigment has no dry- 
ing quality different from that of any iron-oxide pigment. The 

145 



146 BESSEMER PAINT. 

quick dr3riiig of the coating is also aided by atmospheric exposure, but 
when the slags are pulverized, these features will not protect any 
associated pigment or substance from the action of the atmosphere. 
All slag pigments are electro-negative to the metallic base pigments, 
and to all of the metals that constitute a part of their composition, 
also to the metallic surface that is coated with them. 

Insulating qualities are claimed for Bessemer paint; but other 
paints free from metallic oxides also have this quality. The insulat- 
ing qualities of any paint are due to the vehicle more than to the pig- 
ment, with the single exception of india-rubber. In any case, a paint 
coating cannot resist electrical currents of high potential; to mod- 
erate or low potential the insulation would be more or less resistant 
according to the amount of resinous matter in the vehicle. In 
this respect Bessemer paint, containing as it does a small amount 
of fossil resin, would be better than a paint containing none. 

Common resin or resin oil should not be substituted for copal; 
they are not desirable elements in a paint, as they dry hard and 
crack the coating or cause it to crumble and rub off after a short 
exposure in the open air or sunlight, and they promote corrosion. (See 
Paint Tests, Chapter XXIX.) 

Pulverized mineral wool has been proposed for a pigment. It is 
a furnace slag riven when in a molten state by a current of steam. 
But merely pulverizing it imparts no protective value to it for a 
pigment. It is acid in reaction, electro-negative in character, and 
when used for covering steam-pipes or other ferric bodies, on becoming 
damp is a virulent agent for promoting corrosion. A sample of min- 
eral wool analyzed by Prof. Elgleston, of Columbia University, gave 
the following result: 

Soluble Insoluble 

Substances. Per cent. in Water. in Water. 

Water 1.08 1.08 

Potash 0.19 0.19 

Soda 1.75 1.75 

Magnesia 19.82 0.12 19.70 

Lime 26.66 1.61 24.95 

Sesquioxide of iron 0.64 .... 0.64 

Alumina 7.84 7.84 

Silica 38.97 38.97 

Sulphur (mostly as a sulphide, 

probably, of calcium) 2.64 0.32 2.32 

99.49 5.07 94.42 



BESSEMER PAINT. 147 

The composition of this slag is not much different from the preceding 
analysis of blast-furnace slag, and with it may be taken as representa- 
tive of this class of substances. 

No analysis of the Bessemer pigment is given by the manufacturers 
of the paint. It is, however, supposed to be a tetra-basic phosphate 
of lime, containing about 20 per cent of phosphoric acid and 50 per 
cent of lime, associated with other mineral and metallic substances 
in Bessemer iron ores. Some of these substances are partially con- 
sumed in the working of the furnace, and the balance fluxed off as 
Bessemer basic slag. 

Bessemer basic process steel was made by the Pottstown Iron 
Co., of Pottstown, Pa., for. a few years previous to 1893, and the pul- 
verized slag was sold for a fertilizer. Since 1893 the basic process 
for making steel has been suspended in America, and the slag is now 
procured from Germany. 



Analysis of Bessemer Converter Basic Slao. 

Phosphoric acid 21 .37 per cent. 

Lime 45.26 " " 

Ironoxide 12.00" « J Equal to 8.40 per cent of 

( metaluc iron. 

Silica 5.10 " " 

Magneaa 5.90 " " 

Alumina. 4.01 " " 

Soda and potash 0.80 " " 

Manganese oxide 5 .56 " " 



100.00 



« <* 



An analysis of blast-furnace slags, the mean of 2000 samples 
from furnaces working on gray forge pig iron, is given for a comparison 
with the Bessemer converter slag. 

Silica 43 .07 per cent. 

Lime 28.70 " " 

Alumina 14.83 " " 

Ironoxide 2.83" " « Equal to 1 . 98 per cent of 

( metallic iron. 

Peroxide of magnanese 1 .37 " " 

Magnesia 6.46 " " 

Potash 1.84 " " 

Calcium 1.01 " *' 

Sulphur 0.89 " " 

100.00 " " 

Specific gravity 2.8 to 3.2. Fracture vitreous, similar to broken 



148 BESSEMER PAINT. 

earthenware. Color dark gray, tending to the darker or brownish 
shades. 

Blast-furnace slags are acid in reaction, while Bessemer slag is 
basic or neutral. Both are pyrogenic bodies \inaflfected by heat or 
sunlight, and neither is oxidized by the atmosphere. 

A German chemist * gives the analysis of Bessemer paint, as known 
to the trade in Germany, as follows: **The pigment contains baryta, 
alumina, iron oxide, lime, silica, zinc oxide, sulphuric acid, carbon 
dioxide, and phosphoric acid." No definite percentages of these 
substances are given in the analysis, nor any mention whether they 
were separate constituents of the paint assembled in the process 
of grinding and mixing, or that any number of them were foimd com- 
bined together as a single pigment. "Graphite or other carbon is 
used as coloring matter, and linseed varnish as the vehicle, turpentine 
constituting the drier. The presumptive constitution is, therefore, 
Hthopone, or silicious calamine ore, containing baryta and chalk, 
together with graphite or other form of carbon, and linseed varnish 
(with probably turpentine as a drier). When treated with hydro- 
chloric acid it disengages sulphureted hydrogen." 

There are many formulae for compound pigment paints in this 
country, each of some declared excellence by the manufacturers of 
them, if not by the users; but it is difficult to select one that for the 
varied composition will equal this German product. Whatever may 
be the composition of the American brand of "Bessemer paint," it 
appears, from the above description, to be wholly unlike that of its 
German namesake, and is certainly superior to it for a ferric paint. 

* "Andes' Iron Corrosion.'* 



CHAPTER XV. 



NATURAL-BOCK HYDRAULIC CEMENT. 

Hydraulic-cement coatings, either in the form of a plaster coat 
laid on by a trowel or as a wash or brush coating, have not been much 
used for the protection of ferric structures as a substitute for paints, 
though its use as a protection from corrosion of iron embedded in 
masonry is common and its value for this purpose imder certain condi- 
tions is recognized- 
Hydraulic cement made from the ground mineral varies greatly 
in quality, its general composition after calcination, that makes it 
caustic and anhydrous, being: 

Lime 50 to 80 per cent i 

Clay 25 " 40 " " >â–  Specific gravity, 1.6 to 1.6. 

Iron oxide.. 3 " 14 " " J 

Silica, sand, magnesia, sulphur, and many metallic oxides are also 
present in some amounts in many varieties of the hydraulic mineral, 
all of which aflfect the quality of the cement unfavorably when it is 
used for a mortar, and are more objectionable when it is to be used 
for a coating on ferric bodies. 

The adulteration of mineral hydraulic cement is generally, from the 
same class of minerals of inferior quality, with free sand, silica, and 
iron ores containing sulphur in the form of sulphides, and all are imper- 
fectly roasted and pulverized. 

Their setting quality and strength are very irregular and uncertain 

whatever their trade name, or the manufacturer's report of the large 

quantities sold. 

Analysis of an average brand of the so-caUed Portland cement: 

149 



150 NATURAL HYDRAULIC CEMENT. 

Silica (SiOO 20.36 per cent 

Lime (CaO) 61.90 " " 

Alumina (A1,0,) 7.26 " " 

Magnesia (Mg.O) 3.10 " " 

Iron oxide (FejO,) 3 24 " " 

Insoluble residue (clay and sand) .. . 0.44 " " 

Sulphur anhydride . . (SO,) 1.36 " " 

Carbonic anhydride .(CO,) 0.33 " " 

Water (H,0) 1.97 " " 

Soda ^^"»^U and loss 004 " " 

Potash (K,0,)f*°'^^°^- "•"* 



100.00 



« It 



Tensile strength of the neat cement at the end of 7 days equals 
613 pounds; at the end of 28 days, 800 pounds; with one part of 
cement and three parts of sand for the same periods, 228 and 360 
pounds. 

The composition of a natural-rock hydraulic cement from Chatta- 
nooga, Tenn., is 

Silica 22. 17 per cent. 

Lime 65.68 " " 

Alumina 8.20 " " 

Magnesia 1 .45 " " 

Oxide of iron 2.50 " " 



100.00 



H it 



It is a natural Portland cement similar in character, but superior 
to that found at Boulogne, France. The Chattanooga deposit is in 
the form of layers, and is over 50 feet thick. After calcination at 
a white heat, the following are the average results from a number of 
tests of briquettes. 

After an exposure of one hour in the air and 23 hours in water, 
the tensile strength averaged 235 pounds. After one day in the air 
and 6 days in water, 623 pounds. After one day in the air and 27 
days in water, 797 pounds. It is the strongest natural-rock cement 
in the world. 

Portland cements are conmionly made from a dual combination 
of the following substances: 

1. Limestone with from 18 to 20 per cent of clay. 



SLAQ HYDRAULIC CEMEST. 151 

2. Chalk and clay. Marl and clay. 

3. Clay-bearing limestone (argillaceous limestone) with clay, shale, 
or slag sand. 

These substances are pulverized and mixed in some proportions 
that vary with the different manufacturers. The mixture is then 
calcined to nearly the point of fusion, or to actual fusion, forming a 
cinder which is finely pulverized and called Portland cement. 

Furnace slag, the waste product from blast-furnaces (see Analy- 
sis, Chapter XVI), is also used as the base of Portland cement. 

The slag is heated in mass and quenched in water to granulate 
it, making slag sand; then dried and mixed with about one-fourth 
part of slacked lime, and hen finely pulverized to form the cement. 
Slag sand usually contains from 0.5 to 1.5 per cent of the sulphide of 
iron, that has a tendency to oxidize on exposure to the air, which 
action is destructive to cement in above-ground situations. Slag 
cements are from 6 to 8 hours in setting, as against 1 to 3 hours for 
the American brands of mineral cements mider test of a one-pound 
needle. 

The color of slag cement is a delicate lilac or almost white. If 
the slag sand has been roasted prior to pulverizing, it is generally 
of a dark color similar to the ordinary brands of dark Portland cement. 
If a greenish color is present in the cement, it denotes the presence of 
a large quantity of the sulphide of iron. This greenish color is also 
found in some brands of the ordinary Portland cement, where the 
substances from which it is made contain iron sulphide, and when 
there has been a deficiency of heat in the oxidizing flame of the kiln. 

Slag cements, when mixed and exposed to the air, must be well 
covered, else they will crack, though they harden under water without 
swelling or any material change in volume. They are completely 
hydrated during the process of manufacture, do not require aging, 
and do not deteriorate in storage. 

The character of furnace slag and all the processes of its man- 
ufacture into cement require as close attention to secure a reliable 
product as is required for the mineral or Portland cements. The 
engineer should select a cement for a wash coating for ferric surfaces, 
or for a mortar, by its properties, not its name, and should require 
the standard he desires to reach to be met by systematic and rigid 
tests of every invoice of the cement, and many tests from each in- 
voice, during their mixing and application, regardless of the names 
on the package 



152 SLAG HYDRAULIC CEMENTS. 

The nature of the cement has much to do with its e£Fectiveness. 
The quick-setting cements require the most care in their applica- 
tion, and are generally the best for ship work. If the thin cement 
wash once sets in the bucket it will not again set if stirred up — it is 
then useless. Constant stirring of the paste is necessary, as fifteen 
to twenty minutes after mixing suffices for it to set if not kept con- 
stantly stirred. 

The Portland cements set slower than the American or natural 
mineral cements. Quicklime is sometimes added to delay the setting, 
but renders the cement more caustic and destroys the protective 
qualities of the vehicle in the underlying paint, and opens the way 
for moisture in the cement to reach the metal, and all of the coat- 
ings soon peel or flake off either from the corrosion of the iron or 
by the destruction of the bond between it and the paint. 

Cement coatings, unless spread where they are freely exposed to 
a circulation of the air, are damp, and being porous, are not proof 
against the penetration of gases or liquids. If by accident they are 
exposed to the action of any copper scales, scrap-metal, or water 
charged with acid or alkaline substances, the soluble salts of copper 
thus formed will penetrate the coating, deposit the copper upon the 
iron or steel surfaces, set up a galvanic action, and corrode the 
metals beneath the cement coating. 

The hardness and rigidity of cement coatings render them liable 
to flake off the metallic surfaces under strains due to changes in tem- 
perature of the metal that the cement cannot follow. Such places, 
though of minor extent, are generally inaccessible, and are quickly 
corroded; this action being hastened by the difference in potential 
between the exposed metal and the cement coating, even under ordi- 
nary atmospheric conditions. It is more rapidly developed if acidu- 
lated solutions or vapors are present, as they nearly always are aboard 
ships. 

All of these disadvantages in the use of cement can with proper 
care be in a great measure lessened, if not altogether avoided. Cement 
coatings are in many cases the only protection that can be used to 
prevent corrosion, or to arrest it, even where it has progressed to an 
extreme or dangerous point. 

Wrought-iron and steel inlet, stand-pipes, towers, and other parts 
of water-works metal, are subject to virulent corrosion. The usual 
oil-paint coatings are so soon destroyed by the effects of water in 



HYDRAULIC-CEMENT COATINGS. 153 

motion as to be nearly useless; and this is especially the case if the 
paint is iron oxide or if it has been spread over millnscale. 

The collapse of many stand-pipes shows in nearly every case that 
corrosion was the principal cause of their failure, as its progress at 
one or more places had reached such a degree that it only needed 
a small extraneous disturbance to wreck the structure. A case in 
point is that of the 60-inch-diameter wrought-iron inlet-pipes 300 
feet long, and the lower sections of the stand-pipe of a large water- 
works erected in 1860 and in use only a few years, when corrosion 
had developed upon the inner surfaces of the pipes in so many large 
spots and blisters, in such an irregular manner, that the engineer 
reported "that over one-half of the strength of the pipes to resist 
external pressure had been destroyed. Parts of the pipes were un- 
affected, the mill-scale and shop-marks being in place, while nearly 
one-half of them presented an appearance of being inoculated with 
poison." 

The inlet-pipes being buried in river-silt containing a large amount 
of clay, were comparatively unaffected, though below the water- 
level; but it was still water, not subject to motion like the suction 
and force sides of the pipes. 

The inside surfaces were scraped clean as possible, and then coated 
with one coat of neat hydraulic cement from i to J inch thick, laid on 
by a trowel by house-plasterers. 

The water-tower was wrecked by a tornado in 1890, and all the 
pipes were found free from rust in any degree, and probably would 
have lasted indefinitely. 

Imported Portland cement was used on one part of the pipes and 
Louisville cement for the rest. The former set slowly and had an 
indifferent adhesion to the iron. The Louisville cement set promptly 
and was easier to apply. 

Many other instances of the successful use of hydraulic cement in 
similar situations could be cited. The quality of the cement, the 
manner of mixing and applying it, and the personal equation of both 
the engineer and the employ^, are factors for success. Failures of 
cement coatings are more frequent than successes for the reason that 
in their application one or more of these requirements have been 
neglected.* Prof. J. M. Porter of Lafayette CJollege divided a sam- 

♦'* Notes on Oment Masonry." By I. N. Knapp, Gas Engineers' Pro- 



154 HYDRAULIC -CEMENT COATINGS, 

pie of a well-known and reliable Portland cement into nine parts, and 
sent each to a different physical laboratory with the request that tests 
be made of it in a mortar, one part cement to three of sand, according 
to the rules recommended by the committee of the American Society 
of Civil Engineers. The resulting average strengths reported from each 
of the nine samples were as follows: 75, 102, 114, 136, 153, 163, 176, 
225, 247 pounds tensile strength per square inch. Average for all 
the samples was 153 poimds, and the lowest strength was but 30 per 
cent of the highest. 

If these results produced by experienced men in permanent labo- 
ratories vary so much with the same cement, what is to be expected 
from the inexperienced and careless laborers who are generally em- 
ployed to mix and apply concrete, mortar, or cement coatings ? 

Neat hydraulic-cement coatings crack, they set so rapidly that 
there is always a probability of their setting before the workmen can 
spread them, and the tendency of the workmen to ^^ knock them up^^ 
when they indicate setting or have set, instead of mixing a fresh batch, 
is almost irresistible, the result being a coating of very uncertain 
character, — streaks of firm and close-clinging cement alternating with 
those of dead cement, that readily yield to a slight change in the 
temperature of the metal or covered surface, or to a slight mechani- 
cal injury or a frost. A strong heat from the sun also causes them 
to flake off. 

Bad milling and insufficient burning are a frequent cause of poor 
cement, also an excess of magnesia in the limestone or added adulter- 
ant. Magnesia causes a chemical change or disintegrating action in 
the cement when wetted in the mixing. Free, natural sulphate of 
lime is a dangerous impurity. A low specific gravity and a dark- 
brown color are indicative of poor burning, and are easily detected. 

For the protection of iron or steel beams or grillage, laid as the 
foundation or structure work below the water-line to be embedded in 
cement concrete, the metal should be bright and free from mill-scale, 
which is an electro-negative element, and with the moisture present, 
is certain to produce a galvanic couple with the iron it covers and 
promptly start the corrosion, that will proceed uninterruptedly so 
long as any metal is left for it to act upon. Every atom of the red 
rust as it forms, being also electro-negative, increases the galvanic 

ceedings American Gas Association, New York Meeting, Oct. 16, 1902. Ameri- 
can Gas Light Journal, Nov. 10, 1902, pp. 665-673. 



POROSITY OF HYDRAULIC CEMENT, 165 

energy on the remaining metal. The rate of this corrosion will prob- 
ably reduce the beams in less than one hundred years to a condition 
where their strength to sustain any incumbent load will be no greater 
than an equal quantity of tan-bark. 

No paint coating on the metal can resist the galvanic action, and 
it should be applied solely to prevent any slight corrosion that may 
occur from the time the metal is cleaned until it is laid in situ. 

A lampblack and oil, or a graphite paint are good anti-corrosive 
coverings, but best of all coatings for these situations is a refined 
bitumen and dead-oil mixture applied hot. Oxide of iron, or any 
other coating containing electro-negative substances that induce 
corrosion under atmospheric conditions, will only add to the strength 
of the galvanic couple, by bringing their oxidizable elements into the 
field. The nature of the soil in which the metal is directly in con- 
tact will also contribute to the corrosive action through the galvanic 
couple, blue clay or solid rock only excepted. 

Concrete, as generally laid, is very porous, and is seldom so pro* 
portioned, mixed, or rammed in place as to enable it to fill the voids 
in its mass, and capillary action will enable the moisture and soil acids 
to reach the metal. In all such foundation work the cement should 
be of the best quality, free from sulphur elements, the filler should 
be of small size, and the sand absolutely free from salt or sea-sand 
in order to minimize the dangers of corrosion. These precautions are 
seldom if ever taken, even in part, much less as a whole. The neglect 
of these particulars will soon be apparent in the failure of many an 
important structure whose life will be measured by a few decades 
instead of centuries. That the corrosion in these cases is out of 
sight and mind and irreparable will be the more aggravating. 

Porosity of Hydraulic Cement. 

In a general way, engineers and architects are inclined to blindly 
trust hydraulic cement in many locations where in parallel cases it 
has failed. The quality of a cement suitable for a concrete block 
would not be advisable in a wash coating for a wall or to bed an anchor 
bar. AD hydraulic cements are not only porous but permeable.* 
Quay walls laid in beton blocks, composed of about 440 pounds of 
Portland cement to each cubic meter of sharp clean sand and mixed 

♦ Excerpts from a Report of M. M. Leon Durand Claye, Engr. in Chief of 
Bridges and Roads, Paris. " Annales des Fonts et Chau»5e8/' Mav. 18S8. 



156 HYDRAULIC-CEMENT COATINGS, ADHESION OF, 

with from 300 to 440 pounds of water, the walls being surmounted 
by ashlar masonry, were disintegrated in less than a year by the 
action of sea-water. There was a movement or change in the char- 
acter of the beton, even when 660 to 880 pounds of Portland cement 
per cubic meter of sand was used. In parts of the work where they 
had not been exposed to the action of sea-water, the beton of all pro* 
portions of cement and sand were not only very porous but very 
permeable. Under a head of about three feet, the permeability was 
indicated by a rapid fall of the water in the vessel where the beton 
block was under test. The permeability of the cement vxis in all cases 
accompanied by a disintegrating effect in the beton. The disintegration 
was found to be due to the formation of perceptible quantities of the 
sulphate of lime by the action of the sea-water on the Portland cement. 
The sulphate of lime, when formed in the mass of concrete, solidified 
more or less completely in crystals of such a nature as to develop con- 
siderable molecular activity. Some of the beton cement analyzed 
.75 to .80 per cent of sulphuric acid. 

The greater the amount of water used to mix the cement the greater 
was the permeability and porosity of the concrete, even with all pro- 
portions of the cement and sand. Mortar made with 7 per cent of 
water was very permeable, and this increased perceptibly as the per- 
centage was increased to ten and eleven. In all cases where sea- 
water instead of fresh water was used to gauge the cement, the bad 
effects in the mortar were at once apparent. 

Prof. Bauschinger's experiments showed that the adherence of a 
first-class Portland cement to a bright wrought-iron floor beam was 
625 pounds per square inch; that mixtures of two parts of fine, 
sharp bank sand to one of cement reduced the adhesion to about 70 
per cent of the above value. In mixtures of three parts of sand to 
one of cement the adhesion value was less than 50 per cent. That 
with each increase in the percentage of sand from the above amounts, 
the reduction in strength and adhesion was very rapid. The quality 
of the cement had a great effect upon the adhesion value. In the 
commercial cements usually provided for contract concrete, the adhe- 
sion was frequently only 20 per cent of that given above. 

Bloxam's "Chemistry," edition 1895, pp. 376, 377, states "that 
the ordinary corrosion of iron is accomplished only in the presenc>e of 
moisture, air, and COj. If any of these substances are absent the 
corrosion cannot take place. The reactions are: 

Fe+H20+C02=FeC03-fH3 (a) 



HYDRAULIC-CEMENT COATINGS, DEFECTS OF, 157 

The FeCOg is dissolved by the carbonic acid present, and the solution 
absorbs oxygen from the atmosphere, in accordance with 

2reC08+0=reA+2C02 (0) 

The FcjOj combines with the moisture and is deposited as 
2Fe20j.3H20, or ordinary iron rust. Iron in its ordinary state is 
not affected in perfectly dry air, and it will not rust in water con- 
taining a free alkali or alkaline earth or an alkaline carbonate, 
because the affinity of these alkaline substances for any add is 
greater than that of iron, so that they would neutralize the acid 
before it had time to attack the iron." 

This neutralizing action, however, would only be effective for a 
short time or until the alkaline substance became saturated with the 
acid element. There are no locations where concrete is used or cement 
coatings applied to iron for its protection from corrosion where reviv- 
ing the saturated alkaline substance is possible. It is therefore only 
necessary to have a limited amount of some acid present with air and 
moisture to cause the ultimate destruction of a large amount of iron, 
because the CO, or other acids present never become fixed, but are 
always active, passing from molecule to molecule, as long as there is 
any free metal for them to attack. 

It is proposed to increase the safeguard afforded by alkaline sub- 
stances to delay corrosion by mixing the concrete, mortar, or wash 
coating with whitewash instead of plain water. The small amount 
of lime thus added to the cement does not materially detract from its 
strength. 

Slag cements, because of the sulphides present, should be avoided 
for use in concrete or any coatings in contact with iron. It is hardly 
possible to assemble them with an amount of any alkaline substance 
that will permanently neutralize the acid element present in their 
composition, aggravated in nearly every instance by the porous nature 
of all concrete constructions caused by deficient ramming to fill the 
voids occupied by the enclosed air, also by the surplus of water used 
in mixing. 

Even a cement free from the sulphur element, if mixed with a 
small quantity of cinder, or if laid in soil containing cinders or pyrites, 
will absorb the acid and collect it in dangerous amounts in the voids 
of the concrete. Once there, it will ultimately reach the metal and 
cause the failure of the grillages by the columns or other superincum- 
bent load punching through the foundations. These conditions are 



158 CORROSION OF STEEL IN CONCRETE. 

further aggravated by the fact that nearly all ground-water is charged 
to some extent with saline or sulphur elements or both, that would 
soon saturate any alkaline substance present in the cement. When 
this point is reached corrosion of the grillage will inevitably ensue 
even if the imposed columns show no evidence of its action. 

Grillage ironwork has been removed from concrete foundations 
laid only five years and found to be corroded J inch or more over its 
whole surface. The thickness of grillage beams is seldom ^ inch, so 
that thirty or fifty years will practically limit the safety of many of 
the modem steel skeleton structures. 

The protection afforded to steel by Portland cement has been sub- 
jected to experiment by Prof. Charles L. Norton.* "Two brands of 
American cement were selected, tested chemically and physically and 
found to be good. A sharp, clean bank sand and fragments of trap- 
rock and flint were thoroughly washed and used for the concrete. 
The cinders were washed and dried; they tested distinctly alkaline 
with a small amount of sulphur. All the ingredients were mixed dry 
in every case, and when wet with a minimum amount of water were 
tamped until they flushed. 

*' Briquettes were made in duplicate with both cements, viz., neat 
cement, one part to three of sand, one part to five of broken stone; 
cement one part, two of sand, and five of stone; cement one part, sand 
two parts, and five of cinders. Specimens of mild-steel rods 6"+i" 
diameter, mild sheet-steel plates 6"+l"-f^" thick, and strips of ex- 
panded metal 6"+l" were all cleaned bright. All three pieces were 
put into each briquette and were enclosed in separate tin boxes, which 
also contained a specimen of each metal unprotected. One-half of 
the briquettes were set in water for one day and the rest for seven 
days before sealing them up tight. One-quarter of the boxes were 
then subjected to each of the following exposures. To an atmos- 
phere of steam, air and carbon dioxide; to air and steam; to air and 
carbon dioxide, and the other samples set upon a table in a room with 
no special care as to their temperature or dryness. 

''At the end of three weeks the briquettes were cut open and the 
steel examined and compared with the specimens which had lain 
unprotected in each of the tin boxes. 

" The specimens covered with neat cement were as bright as when 
placed in the briquette, the cement had prevented any trace of corro- 

* Engineer in charge of the Insurance Engineering Experiment Station, 
31 Milk Street, Boston, Mass. Excerpts from Third Report, 1902. 



CORROSION OF STEEL IN CEMENT. 159 

sion, while the unprotected samples consisted of more rust than steel. 
Of the remaining specimens hardly one had escaped serious corrosion. 
The location of the rust spot was invariably coincident with either a 
void in the concrete or a badly rusted cinder. Rust had as usual pro- 
duced rust. 

** In the more porous mixtures the steel was spotted with alternate 
bright and rusty areas, each clearly defined. In both the solid and 
porous cinder concrete many rust spots were found, except where the 
concrete had been mixed very wet, in which case the watery cement had 
coaled nearly the whole of the steel like a paint and protected it. 

"Some briquettes made of finely ground cinders and cement in 
varying proportions up to one of cement to ten of cinders and exposed 
to moisture and carbonic acid showed how effectually the presence of 
the cement prevented rusting, provided there were no cracks or crev- 
ices or distinct voids. The corrosion found in cinder cement appeared 
to be mainly due to tlie iron oxide ia- the cinders and not to the 
sulphur. Cinder concrete, well rammed when wet to fill the voids, 
is about as effective as stone concrete in protecting steel." 

These latter conclusions would depend greatly upon the absence 
or low percentage of iron oxide and sulphur in the cinder. To render 
these dements inert to iron, there must be enough free alkaUne sub- 
stance in the cement to saturate the acids without disturbing the 
general composition of the cement as a binding element. 

If the metal is painted before the application of the concrete, what- 
ever its composition, the continuous void left over the whole surface 
of the metal by the decay of the paint is the best possible condition 
for inaugurating corrosion. Air and moisture will find ready access 
to this void, also to the voids left by building bricks and terra-cotta 
blocks, the porous nature of which are favorable to the condensation 
and absorption of moisture and atmospheric gases, that are more 
highly charged with corrosive elements in cities, tunnels, subways, and 
other locations where the use of structural steel work is in most de- 
mand. 

How far the protection of ferric foundations, either near or below 

the water-line in the many structures already built, or in progress m 
all parts of the world, has been considered by their architects and 
engineers time only will reveal. For those proposed, like the miles 
of rapid transit and railway tunnels, a great portion of which wUl be 
carried through ocean silt or salt-marsh mud and exposed to the most 



160 CEMENT FOR TUNNEL WORK. 

virulent form of corrosion, some more positive and effectual means of 
protection from corrosion must be employed than has ever been 
adopted. No wash or trowel coating of cement, good or bad, or 
applied in mass, will avail for but a short period to protect the metal 
that these structures must rely upon for a great part of their strength. 

The hardness and inelastic character of cement or mortar coatings 
will cause them to crack under the vibrations inevitable to all rail- 
way structures; and while resisting water in mass, they will absorb 
moisture sufficient to be always damp and in that condition are of 
the least strength. 

The wires of the anchorage ends of the cables of the Niagara Falls 
suspension bridge were opened for a short distance where they entered 
the anchorage pits. These ends were embedded in hydraulic cement, 
and at the end of forty years many of them had become so corroded 
that the strength of the structure was seriously impaired. The cor- 
roded strands were replaced by new wires, and the top part of the 
anchorages opened to allow the cement work to dry out and remain 
dry. In this case and with all ferric material embedded in concrete, 
the caustic action of the usual make of cement, whether damp or 
wet, will furnish the carbonic acid necessary to destroy any linseed- 
oil coating or paint that covers them and induce corrosion. The 
subsequent drying out of the cement coverings only delays for a 
short time the ultimate destruction of the metal. 

Iron anchor bars and chains embedded in concrete below the water- 
line for 100 and 200 years were free from rust when removed. Fresh- 
water immersion, no access of air, no acid element nor iron oxide or 
calcium sulphate, also no voids in the cement was the secret of 
their perfect condition. 

The present method of constructing buildings w^holly or in part of 
steel framing and concrete, avoiding the use of brick and stone ma- 
sonry as far as possible, is causing a great deal of anxiety among archi- 
tects and engineers as to the future state of the metal so embedded. 
That metal needs some additional protection from the caustic action 
of the impure cements too frequently employed, also from the quick- 
lime mortar, beyond the usual coat of paint, is recognized. 

At a late meeting of the English Architectural Association, Mr. 
H. Humphrey gave as the result of his experience that metal buried 
in concrete containing furnace cinders or coke breeze, should be 
coated with Dr. Angus Smith's anti-corrosive compound, or some 



CINDER IN CEMENT COATINGS. 161 

other compound containing pitch and sand; that some samples of 
cinder concrete analyzed as high as three-fourths of one per cent of 
sulphuric acid. A case was cited by another member of the asso- 
ciation where a hot-water pipe laid in cinder concrete was rotted 
away in a very short time. 

The cinder concrete used in the floors of the steel-frame sky- 
scrapers in New York City invariably shows the presence of sul- 
phuric acid strong enough to redden litmus-paper. 

Gas-pipes embedded in plaster of Paris (gypsum) have been found 
to be completely corroded in a few yeara. The use of gypsum in 
cement to hasten its setting is detrimental. Gypsum is soluble to 
some extent in water, besides it contains water from its hydration, 
which absorbs carbonic acid from the air that quickly causes corro- 
sion. The rust so formed absorbs moisture and carbonic acid and 
further hastens the corrosion. 

The screwed ends of all pipes are invariably attacked. They are 
of bright metal only about ^ inch thick and seldom, if ever, have 
even a brush coating of any paint to protect them when put up or 
left in place. Galvanizing the fittings and body of the pipes does 
not protect the screwed ends; the corrosion at these points is only 
hastened by the galvanizing. 

The effect of corrosion upon the floor beams and other structural 
parts used in modem architectural work has been the subject of dis- 
cussion by the American Society of Architects, the concensus of their 
opinion being expressed by one of the prominent members as follows: 

"With regard to the strength of the steel-cage constructions, both 
as to wind strain and other disturbing strains, there is no question. 
All objections arising from these points have been overcome, but un- 
less exceptional care is taken in the construction to protect the steel 
cage, particularly at its joints, from corrosion, this class of buildings 
will not be permanently safe. It is perfectly feasible, with great 
care, to protect the steel frames from corrosion, but I am convinced 
that many high buildings have been put up in this country where the 
proper care in this respect has not been taken nor the necessary 
preventives against corrosion applied." 

On the I2.inch steel I beams carrying the sidewalks around the 
Pabst Hotel in New York City, and that were removed after being in 
place less than six years, corrosion was in active progress. The 
space beneath the beams was used as a caf6, always dry and heated. 



162 CEMENT COATINGS FOR MARINE WORK. 

Wherever the brickwork came in contact with the beams in all of the 
stories, the paint was dead and corrosion established. This was par- 
ticularly noticeable in many portions of the beams where the usual 
top dressing of coal cinders had been laid to level up the arches form- 
ing the foundation for the artificial stone sidewalk. The rivets that 
held the comer angle-irons to the beams were nearly all loose from 
the corrosion around their heads or points and had lost their set or 
draw. 

In marine work, hydraulic cement is used almost exclusively as a 
brush coating on the inside surfaces of the ship's frames and plating 
.in the lower holds of the vessel where the metal is exposed to the 
action of bilge-water, alkaline and acid solutions from acids, and 
leakage from the cargo liquids. One or more wash coatings of cement 
are applied over the red-lead, black-vamish, or other oil-paint coating 
laid on during the construction of the ship, and that generally serve 
to protect the metal during this period. The coatings in the lower 
part of a ship are damp by reason of the confined saturated sea air, 
but the cement (if good) forms a close, clinging coating that seldom 
fails unless by mechanical injury or improper mixing or application, 
and is easily repaired if injured. 

The confined spaces aboard a ship almost preclude the use of an 
oil paint or varnish, however quick drying it may be, without the use of 
forced ventilation to provide the oxygen necessary in the drying of 
paint. Such ventilation is practically impossible in a ship at sea, 
or in most cases in dry dock. 

Cement for rendering, with the object of making brickwork water- 
tight, should be mixed with an equal part of absolutely clean sand free 
from salt or sea-sand. Cement for any use should be carefully turned 
over by the shovel and exposed to the air before being mixed or 
wetted. 

Brickwork is one of the worst surfaces to hold a paint, good or 
poor. A hard-burnt brick will absorb 8 ounces of water, a salmon 
brick nearly 11 ounces. Brickwork absorbs most of the oil in the 
paint, leaving the pigment on the surface of the bricks without suffi- 
cient bond to hold it, and it peels in strips. This peeling is hastened 
by the caustic action of the cement or lime mortar and the soluble 
salts in the sand of the mortar, which soon destroys the organic mat- 
ter in the oil. 

The soluble salts in the bricks and mortar often cause a white 
efllorescence to appear on the surface of the wall shortly after being 



PAINTING OF BRICK WALLS. 163 

built. This cannot be brushed off, but it disappears during wet 
weather and returns again when the wall is dry, and is only dissipated 
after many rain-storms. The composition of the efflorescence varies. 
The chlorides of calcium, magnesium, and sodium sulphate and 
oxalate of lime are generally present; all of which are hygroscopic 
and form a germinating place for fungi, one of which, the PenicUium 
cruOaceum, is represented by Fig. 24.* 

Walls a year or more old are less troubled with the efflorescence 
or by the peeling of the paint from the fungus. 



Fia. 24.— Photomicrograph X 600 of Penieilium a 
fungus which maKes calcium oxalate on b 

When walls are freshly laid or plastered the surfaces can be pre- 
pared for painting by applying a solution made of twelve fluid ounces 
of sulphuric acid in a gallon of water and repeating the application 
when the first one appears dry. Allow the coatings to stand for 
a day or two, then rinse off with clear water, and when dry prime 
and paint as usual. This process changes the lime in the mortar and 
cement from a caustic carbonate to a neutral sulphate of lime; also 
produces a uniformly absorbent surface free from spots that are more 
porous than the general surface, or that contain lumps of improperly 
slaked or mixed lime. The surface so prepared takes the paint easily 
and well and does not blister nor peel. 

If plastered surfaces a tew months old be washed with a solution 
Journal ol Chemiaillndattry (London), 



164 PAINTING OF BRICK WALLS. 

of 2 ounces of bicarbonate of ammonia in a gallon of water, as soon 
as dry the oil priming or painting can be done without danger of 
peeling. 

A silicate of soda solution made from equal weights of silicate and 
warm water, and applied with a brush, is also recommended for pre- 
venting the peeling of paint on walls, but for outside exposures it is 
not so effective as the above acid treatment. 

WcUerproofing Bricks and Sandstone. At a recent meeting of the 
Australian Association for the Advancement of Science, Professor 
Liversidge read a paper on the "Waterproofing of Brick and Sand- 
stone with Oils'." Experiments were made with the view of ascertain- 
ing the length of time that brick and sandstone are rendered water- 
proof or protected by oil. The oils used were the three commonest 
and most readily obtainable for such purposes, viz., linseed-oil, boiled 
linseed, and the crude mineral oil known as "blue oil," used for pre- 
serving timber. The weatherings were made upon a flat portion of 
the laboratory roof fairly exposed to the sun and weather. Good, 
sound, machine-made bricks were experimented on. The amount of 
oil and water taken up by the sandstone was very much less than 
that absorbed by the brick, although the area of the sandstone cubes 
was much greater than that exposed by the bricks. Equal amounts 
of raw and boiled oils were absorbed; the blue oil, however, was taken 
up in much greater quantity by both brick and sandstone, but by the 
end of twelve months the whole of the 13J ounces of blue oil had 
apparently evaporated and the brick had returned to its original 
weight. The bricks treated with raw and boiled oils remain unchanged. 
After the second oiling in November, 1890, and exposure for nearly 
four years and two months, they had practically retained all their oil, 
inasmuch as they had not lost weight, and were also nearly impervious 
to water. It was noticeable that the sandstone cubes treated with 
raw and boiled oils returned to their original weights, but did not 
appear to have lost the beneficial effects of the oils, being also 
practically waterproof. 

Portland or other hydraulic cements free from the sulphate of lime, 
when mixed with about 15 to 25 per cent of a red-lead paint, forms a 
tough elastic coating that dries hard enough to resist the action of 
locomotive exhaust steam and cinders on the surfaces of iron beams, 
trusses, and the buckle-plates of low headway bridges. It is also 
damp-resisting to a great degree, and can be coated over with other 
paints with some measure of confidence that they will not peel. 



PAINTING OF BRICK WALLS. 165 

A special damp-resisting paint) known as the "R. I. W." (trade- 
mark), has proven of merit in many instances of its use in very un- 
favorable situations. From a government analysis of it, the com- 
position is approximately 30 per cent of an oil vehicle, 65 per cent 
of refined special bitumen and selected fossil resins, and 5 per cent of 
a carbon pigment. It is laid on or spread like a thin coating of mortar 
on brick masonry or plastered walls. It adheres firmly, becomes very 
tacky, and can be plastered over with cement or lime-mortar coatings 
that adhere firmly. When these plastered coatings are dry they can 
receive an oil paint of any desirable color, unaffected by dampness 
from the walls. 

A grade of the "R. I. W." is also made to apply to damp walls not 
intended to be plastered, also to iron structural work. This is applied 
with a brush, and contains more pigment than the trowel grade. It 
is thoroughly damp-proof, and receives oil-paint coatings without any 
tendency to craze them or to peel. 

A grade of this composition, to be spread with a brush on the inside 
of tanks where acid and alkaline solutions are stored, effectually resists 
the action of these liquids. In chemical works for the protection of 
the ironwork and other metals, it has shown great resisting power. 
A special instance of the waterproofing character of this compound 
to resist the action of running water under a considerable head is on 
the concrete monolithic water-power house on the St. Lawrence River 
at Massena, N. Y., where several thousand square yards each of the 
trowel and brush coatings were applied, and completely corrected the 
porosity and penneability of the cement walls that were seriously en- 
dangering the structure. 

Herr Wm. Cremer, superintendent of the gas-works at Enskirchen, 
Germany, states: "That the ammoniacal liquor from gas-works, even 
in the weakest solutions, detrimentally affects the cement and tank- 
walls exposed to its action. Coating the surfaces thus exposed with 
liquid glass (tungstate of soda) protects the cement, also renders the 
surfaces quite leakage-proof even in very old work." 



CHAPTER XVI. 

BOWBR-BARFF COATINGS. 

^* Bower-Barff " is the name given to the rustless coating fonned 
upon cast iron, wrought iron, and steel, when exposed to a low red 
heat in special ovens, furnaces, or retorts, and subjected to the action 
of superheated steam, carbonic-oxide gas from coal-fires or gas-pro- 
ducers, hydrocarbon and hydrogen gas alternately or in combina- 
tion, according to the several processes invented by Bower-Barff, 
Wells, Gesner, and other inventors. 

The original inventor of the rustless iron coating was Prof. Fred- 
erick S. Barflf, of Kilbum, England, who published an account of his 
process in 1876, and read a paper describing it before the Society of 
Arts, London; but the process did not prove commercially successful 
on account of its high cost and the difficulty of obtaining uniformity 
in results. 

Messrs. George and Anthony Bower, of St. Neats, England, im- 
proved the process of Prof. Barflf, and patented it. The right to use it 
in the United States was acquired by Mr. George W. Maynard, of New 
York. The first furnace was erected at the Hecla Architectural Iron 
Works, in Brooklyn, N. Y. 

The next and most important of the improvements in this process 
was invented and patented in 1888 by Mr. W. T. Wells, of Little Ferry, 
N. J., who discovered that red-hot iron, in the presence of mingled 
steam and carbonic oxide, would form the magnetic or black oxide 
(rustless coating) of iron, Fe804, without the intermediate formation 
of the sesquioxide, FejOj (red rust), the reactions being, 3Fe+4H20. 
=Fe304+4H2. This process is the foundation for all subsequent im- 
provements of the process and is applicable to all forms of cast, mal- 
leable, wrought iron, and steel where the surfaces are not to be sub- 
jected to hard friction or wear, such as bending, hammering, chipping, 
or other rough usage. 

The protection of the metal being due to the conversion of the 

surface of the metal to magnetic oxide, and not any material altera- 

166 



BOWER^BARFF COATINGS, 167 

tion of the metal which would weaken it, it follows that any manipu- 
lation that would injure the contmuity of the coating must neces- 
sarily destroy the coating. Wherever the coating is broken the metal 
will rust, though the rust will be localized, and will be greater than the 
same exposure of the metal not coated, owing to the difference in 
potential between the two surfaces. 

These rust-spots seldom spread or raise the adjacent coating, as 
is commonly the case with paint, or enamelled coatings. All drilling, 
fitting, screw-cutting, etc., of the metal should be done before 
it is put into the converting-oven. In riveting, the oxide in the 
immediate neighborhood of the rivets will be broken, and bolting 
together of parts to be connected together must be substituted. In 
work that is riveted up before being coated, the set or draw of the 
rivets will be released by the heat of the furnace. This, in the case of 
light grill, lattice, or fence work, is possibly of small moment, but in 
work subject to the action of liquids or gases it cannot be ignored 
and other methods of joining the pieces must be adopted. Shear- 
ing, flanging, sharp bending, or driving of nails through sheet-iron 
roofing, necessarily exposes the metal, and local corrosion of the in- 
jured part follows. The bite of the vise or pipe-wrench in fitting 
rustless screwed steam- or water-pipes injures the coating unless 
special care and tools are used to prevent the injury. The screwed 
ends of pipes and fittings are injured if the joints are made up dry, 
but with red lead, graphite, or other good pipe-joint cements as lubri- 
cants, they seldom give trouble if moderate care is exercised in the 
work. 

In cast-iron pipe with bell and spigot joints, the lead packing can 
be calked without injury to the "rustless" coating by using the round- 
nose calking-tool, instead of the usual sharp-edged tool that chips the 
coating. Rustless pipe coatings do not appear to draw in the lead 
joint any more than the usual coal-tar dip coatings, from the changes 
in temperature that all pipes are subjected to when buried in the 
ground. 

The mechanical finish of the metal to be coated determines to a 
great extent the mode of treatment. Articles in the rough, from 
which the skin has not been removed, require a longer exposure, 
higher heat, and a more energetic oxidation than those whose sur- 
faces are more or less machined or finished ; the latter requiring lower 
heats. A high heat on a finished surface tends to blister and detach 
the oxide as fast as it forms. Ordinarily, only the rust and mill- 



168 BOWER-BARFF COATINGS. 

scale are removed by scraping and use of steel brushes. Where a 
handsome appearance of the oxidized ware is desired, the surfaces 
must be cleaned by the sand-blast, or by pickling, and the same care 
used to remove all traces of the pickling acid by a warm lime-water 
bath and repeated washing with cold water-jets under pressure, as 
in the case of cleaning the metal for painting (Chapter XXVIII). 
Foundry-sand upon castings, if not removed, bakes in the furnace 
to .a reddish-brown color, producing unsightly spots, but does not 
impair the rustless character of the coating, and unless the coating 
is to serve as a finish, without being painted, the spots are of no 
moment; otherwise the sand must be removed to the clean-scale 
surface before treatment. All blow-holes and other defects in cast- 
ings must be plugged with brass or iron plugs. Lead or other soft 
fillings are detrimental to the action of the furnace in producing a 
reliable or fine-appearing coating, which should be a pleasing blue- 
gray or blue-black color. If the metal is polished before treatment, 
it acquires a lustrous ebony-black finish, very desirable upon certain 
kinds of articles. 

The iron or steel articles treated, owing to the annealing action 
while in the furnace, are permanently expanded about -^ inch per foot, 
for which allowance must be made where this addition will be repeated, 
as in stair-stringers, columns, etc. 

The limit of elasticity of the oxide coating is practically the same 
as that of the metal it covers. The coating adheres firmly under 
tensilC; torsion and compressive strains, until the elastic limit has been 
reached, and no further. 

In Sir Joseph Whitworth's tests of specimens of Bower-Barffed 
wrought iron, submitted to tensile strain, smaD pieces of the oxide 
coating scaled off when the strain reached 28,618 pounds per square 
inch, or beyond the elastic limit, and about one-half of the ultimate 
strength of the specimen. In the case of cast iron, the coating re- 
mained in place uninjured when strained to the point of rup- 
ture. 

Bower-Barffed articles can be heated to temperatiu*es approxi- 
mately 400*' Fahr. and then immersed in cold water without injury. 
They resist the action of sea air, sea water, sulphurous, and other 
gases, ammonia, and all alkaline and organic acids in moderate solu- 
tion, also the caustic action of lime and hydraulic cement either dry 
or damp. They are, however, affected by strong solutions of sul- 
phuric and hydrochloric acids. 



BOWER'BARFF COATINGS, 



169 



A comparative test of the resistance to corrosion of a number of 
protective coatings under different exposures resulted, viz. : * 

Change in Weight op Wrought and Cast Iron with Different Protec- 
tive Coatings and Under Different Conditions, in Pounds per 
Square Foot of Surface per Annum. 

Wrouoht-iron Sheets (No. 23 Gauge, Black). 



Proteotivo CoatingB. 



Bowei^BarfiFed 

Tinned 

Nickel-plated 

Galvaniied 

Barffed 

Black — i.e., unprotected 
Copper-plated 

Averase gain 



Exposed to the weather 
Inland. 



Canada. 



.0 
gain, .002.0 

.0 
gain, .000.4 
.001.0 
.001.3 
.000.2 



•4 



.000.2 



New York 
State. 



gain, .000.3 
.000.1 
.000.5 






gain, .003.1 
•* .022.6 
" .005.0 



.005.1 



ImmerBedin — 


Fresh 
water. 


Sewage. 


.006.7 
.019.4 
.050.4 
.045.9 
.083.9 
.137.0 
.179.0 


.003.6 
.007.1 
.003.1 
.080.5 
.117.0 
.169.0 
.182.0 


.074.6 


.080.3 



Cast-iron Plates. 



Protective Coatings. 



Bower-BarfiFed 

" and paraffined. 

Galvanised 

Tinned 

Nickel-plated 

Copper-plated 

Black — i.e., unprotected. . . . 

Average gain 



Exposed to the weather 
Inland. 



Canada. 



gain, .004.0 
" .000.6 



.0 



, .003.4 
.004.0 
" .006.3 



gain 



.002.9 



New York 
State. 



gain, .003.1 
.001.9 
.0 

gain, .003.1 
.002.5 
.005.0 
.012.0 



It 



Immersed in — 



Fredh 
water. 



.002.1 



gain,.005.5 
.000.2 
.049.1 
.065.5 
.131.7 
.150.8 
.148.3 



Sewage. 



** 

«4 
It 
t« 
«4 
44 



.001.4 
.008.4 
.061.0 
.061.0 
.083.3 
.119.2 
.272.4 



.007.2 



.086.7 



Average 
gain. 



.002.6 
.006.2 
.013.5 
.042.0 
.051.2 
.082.5 
.091.6 

.040 



Average 
gain. 



gain 



.002.8 
.002.8 
.027.5 
.041.1 
.053.5 
.067.8 
.106.6 



.041 



The cost of appl3dng the process must necessarily vary with the 
size, weight, and other characteristics of the article to be treated. 
For builders' hardware and that class of articles called shelf goods, 
domestic articles, etc., the cost is about 5 per cent of the net cost of 
the goods to the manufacturer. Wrought-iron grilling, office rail- 
ings, and the better class of scroll and fancy work cost about two 
cents per pound. Wrought-iron steam and water-pipe is coated for 
about the same expense per pound as is required to paint it. A 
further benefit to this class of articles is, that the inside of the pipe 



♦Trade catalogue. "Rustless Iron and Steel. The Bower-Barff and 
Wells Processes." (Pamphlet Number V. Illustrated.) By the Bower-Barff 
Rustless Iron Company, New York, 1896. 



170 BOWER'BARFF COATINGS, 

receives the same coating as the external parts. Wrought-iron I 
beams, channels, and other shapes entering into building construc- 
tions can be treated very cheaply; the principal expense is first cost 
of the furnace; the actual operating expenses are very small — frac- 
tions of a cent per pound. 

The Iron Column of Delhi* * 

The iron column of Delhi, India (see Frontispiece) ^ is 20 feet high 
above ground, 16 inches in diameter at the base, and 12 inches at the 
top, with an ornate Persian capital 3^ feet in height. The base has 
a Persian inscription of six lines on the western side, symbolizing 
the deeds of the Rajah Dhawa, who reigned in the ninth century b.c. 
Beck, in his "History of Iron," places its erection in the early part 
of the fourth centiuy a.d., but other authorities place it in the ninth 
century B.C., corresponding to the inscription upon it. 

Early excavations to the depth of 26 feet did not reach the bot- 
tom, but subsequently it was found to rest upon forged-iron beams, 
bedded and anchored to the stone foundations. A short distance 
below the ground it is 2 feet 4 inches in diameter, and evidently was 
forged from a large number of wrought-iron blooms. Its estimated 
weight is seventeen tons. 

It stands alone above all other relics, a monument commemora- 
tive of the state of the mechanical arts in prehistoric times, not only 
for its construction and preservation, but its transportation from 
some unknown and evidently far-distant place of manufacture and 
its erection in situ. This would be considered, at the beginning of 
the twentieth century, an exceedingly creditable example of engi- 
neering skill, and it wiU probably remain centuries after most of the 
present-day ferric constructions have crumbled to red rust. 

It is free from corrosion, and while this in a measure may be due 
to the climate of India not being inducive of corrosion, it cannot alone 
be the reason of its protection, for other iron articles, both large and 
small, bear testimony to corrosive effects under the same exposure 
and climate. A reason for its non-corrosion has been given: that in 
the earlier days following its erection it was considered a part of the 
religious duty of every pilgrim to the holy shrine near which it is 
located to climb it, and as the pilgrims annointed their bodies with oil 
as a part of their devotional exercises, that more or less of this oil was 



* Iron Age, August 1, 1895, p. 215. 



IRON COLUMN OF DELHI, 171 

left upon the column and thus protected it. But for the past two 
hundred years or more, so far as known, no such greased-pole gym- 
nastical devotions have been practised, and the coatings of oil, if any 
ever were thus appUed, must have long since been dissipated, as 
they would doubtless have been palm or some other vegetable oil or 
camel's fat, all of a non-siccative nature. 

The ornate capital is as free from corrosion as the shaft of the 
column, so unless the pilgrims climbed this as well as the shaft (as a 
sailor-boy mastheads his ship's truck), and possibly stood on their 
heads as a further sign of exalted zeal, the capital could not have 
received the oil treatment to protect it. The part of the column 
underground surely had no such acrobatic oleaginous distribution, 
and is comparatively as free from corrosion as the part above ground. 
Every indication in the appearance of the column shows that after 
it had been forged and finished, the inscriptions and capital still 
bearing the chisel-marks on the ornaments, it was subjected for its 
entire length to a process identical to that of the modern Bower- 
Barflf process, which has proven to be quite as effective to prevent 
corrosion in this instance as in any of the modem examples of this 
protective method. 



CHAPTER XVII. 

GALVANIZING. ELECTRO-CHEMICAL AND OTHER ANTI-CORROSIVE 

ZINC PROCESSES. 

Galvanizing* 

Galvanizing to protect the surface of large articles, such as enter 
into the construction of railway viaducts, bridges, roofs, and ship- 
work, has not reached the point of appreciation that possibly the 
near future may award to it. Certain fallacies existed for a long time 
as to the relative merits of the dry or molten and the wet or electro- 
lytical methods of galvanizing. The latter was found to bfe too 
costly and slow, and the results obtained were erratic and not satis- 
factory, and soon gave place to the dry or molten-bath processes as in 
practice at the present day; but the difficulty of management in con- 
nection with large baths of molten material, the deterioration of the 
bath, and other mechanical causes limit the process to articles of com- 
paratively small size and weight. 

The electro-deposition of zinc has been subject to many patents, 
and the efforts to introduce it have been lamentable failures in both 
a mechanical and financial sense. Most authorities recommend a 
current density of 18 or 20 amperes per square foot of cathode sur- 
face, and aqueous solutions of zinc sulphate, acetate or chloride, 
ammonia chloride or tartrate, as being the most suitable for deposition. 

Herman's process has been experimented with on a commercial 
scale, the chief feature being the addition of the sulphates of the 
alkalies or alkali earth to a weak solution of zinc phosphate. 

Electrolytes made by adding caustic potash or soda to a suitable 
zinc salt have been found to be unworkable in practice, on account 
of the formation of an insoluble zinc oxide on the surface of the anode 
and the resultant increased electrical resistance; the electrolytes are 

* Excerpts from a paper by the author, presented at the New York meeting 
(December, 1894) of the American Society of Mechanical Engineers, and form- 
ing part of Volume XVI of the Transactions. Also Transactions of the American 
Society of Mechanical Engineers, Vol. XV, 1894, Paper No. 598, pp. 998-1073. 

172 



OALVANIZltiO. 173 

also constantly getting out of order, as more metal is taken out of the 
solution than could possibly be dissolved from the anodes by the 
chemicals set free, on account of this insoluble scale or furring up of 
the anodes, which sometimes reaches ^ inch in thickness. 

To all intents and purposes the deposits obtained from acid solu- 
tions under favorable circumstances are fairly adhesive when great 
care has been exercised to thoroughly scale and clean the surface to be 
coated, and which is found to be the principal difficulty in the appli- 
cation of any electro-chemical process for copper, lead, or tin, as 
well as for zinc, and that renders even the application of paint or 
other brush compounds so futile unless honestly complied with. 
Unfortunately these acid zinc coatings are of a transitory nature, 
their durability being incomparable with hot galvanizing, as the 
deposit is porous and retains some of the acid salts, which cause 
a wasting of the zinc and consequently the rusting of the iron or 
steel. Castings coated with acid zinc, rust comparatively quickly, 
even when the porosity has been reduced by oxidation, aggravated 
no doubt by some of the corroding agents, sal-ammoniac, for in- 
stance, being forced into the pores of the metal. In wrought iron, 
the cinder is porous, and holds the acid, and induces corrosion. 

The relative porosity of zinc coating, applied by different methods, 
is shown by the following micrographs, Figs. 25 and 26, taken from 
The Engineer, September 2S, 1S94. 



Fio. 25.— Zinc coating; applied by Fia. 26. — Deposit from rinc sulphate 
hot iralvanizing process, magnified solution (acid), magnified five diam- 
five (Uameters. etere. 

Other matters of eeriouB moment in the acid electro-zincing proc- 
ess, aside from the slowness of operation, were the uncertain nature, 
thickness, and extent of the coating on articles of irregular shape, 
and the formation of loose dark-colored patches on the work, the 



174 GALVANIZING. 

unhealthy non-metallic look, and want of brilliancy and lustre pre- 
vented engineers and the trade from accepting the process or its 
results except for the commoner articles of use. 

The Cowper-Coles process of electro-zincing articles claims to 
overcome aU these difficulties, and plants are in process of erection 
with a bath of some 14,100 gallons capacity, capable of turning out 
forty tons of light work per week, and in which it is proposed to 
treat the plates of vessels sixty feet in length upon one or both sides, 
and the frames of such vessels as torpedo-boat destroyers and kin- 
dred craft after riveting up. These plates and frames are given a 
thin coating of zinc by this process that appears to be perfectly uni- 
form in character and extent whatever the shape of the piece may be, 
and however numerous the lugs, flanges, mortises, or core-holes. It 
is called *' zinc- flashing"; that is, coating the iron or steel article, 
after pickling and cleaning, mth a thin coat of zinc about one ounce 
per square foot of surface, which resists the inclemency of the weather 
and mechanical injury as well as a thicker coat, and is found to afford 
sufficient protection in most cases, and is adequate protection imtil 
such time as it is ready to receive the usual paint coatings. 

To obviate any tendency of the paint to peel from the zinc sur- 
faces, as it generally manifests a disposition to do, it is recommended 
to coat aU the zinc surfaces, previous to painting them, with the 
foUowing compound: One part chloride of copper, one part nitrate 
of copper, one part sal-ammoniac, dissolved in sixty-one parts water, 
and then add one part commercial hydrochloric acid. When the zinc 
is brushed over with this mixture, it oxidizes the surface, turns black 
and dries in from twelve to twenty-four hours, and may then be 
painted over without danger of peeling. Another and more quickly 
applied coating consists of bichloride of platinum, one part dissolved 
in ten parts distilled water and applied either by a brush or sponge. 
It oxidizes at once, turns black, and resists the weak acids, rain, 
and the elements generally. 

There are r-lso a number of trade-mark, or proprietary, mixtures to 
prevent the peeling of paint applied to zinc. " Uniter," an English com- 
poimd, and "Galvanum," an American paint in light-brown and dark- 
gray colors, are favorably recommended. Carbon and asphaltum 
paint, containing a large percentage of bisulphide of carbon in the 
vehicle, also adheres well to galvanized iron. Its nauseating odor 
and highly inflammable character during its application are strong 
points against its use. 



GALVANIZING. 175 

Galvanized-iron sheets that are corrugated after galvanizing 
corrode more rapidly than uncorrugated sheets. Sharp angles and 
twists in the sheet also corrode quickly. The thin zinc-coating atoms 
are brittle naturally, and are opened to allow moisture to reach the 
metal they cover. This being a more elastic metal, plates coated with 
it do not show the bending effects so strongly, yet they are apparent. 

Double-coated tin, zinc, or terne plates are from two to three times 
more resistant to corrosion than single-coated plates. The second 
coating, like the second coat of paint on a painted surface, fills 
the shrinkage, cracks, and pores in the first coat. Galvanized-iron 
pipes used for gas and water service in the ground have only a life 
of 12 to 15 years, the outside coating of zinc being destroyed by gal- 
vanic action induced by the acid elements in the soil. If the soil 
contains furnace cinders, the corrosion is hastened. The screwed 
ends and other parts of the pipe where the galvanizing has been cut 
away are the parts first corroded. In general all galvanized pipe- 
work is so poorly cleaned from mill-scale and grease prior to galvan- 
izing, that the pipes are less endiu'ing than with a conmion coal-tar 
pitch dip coatmg. 

Zinc surfaces, after a brief exposure to the air, become coated 
"with a thin film of oxide — insoluble in water, which adheres tena- 
ciously, forming a protective coating to the imderlying zinc. So long 
as the zinc surface remains intact, the underlying metal is protected 
from corrosive action, but a mechanical or other injury to the zinc 
coating, that exposes the metal beneath to the presence of moisture, 
causes a very rapid corrosion to be inaugurated, the galvanic action 
being changed from zinc positive to zinc negative, and the iron as 
the positive element in the circuit is corroded instead of the zinc. 

When galvanized iron is inmiersed in a corrosive liquid, the zinc 
is attacked in preference to the iron, provided both the exposed parts 
of the iron and the protected parts are immersed in the liquid. The 
zinc has not the same protective quality when the liquid is sprinkled 
over the surface and remains in isolated drops. Sea air being charged 
with saline matter is very destructive to galvanized surfaces, form- 
ing a soluble chloride by its action. As zinc is one of the metals most 
readily attacked by acids, ordinary galvanized iron is not suitable 
for positions where it is to be much exposed to an atmosphere charged 
with acids sent into the air by some manufactories, or to the sulphuric- 
acid fumes found in the products of combustion of rolling-mills, iron, 
glass, and gas works, etc. ; and yet we see engineers of note covering 



GALVANIZING. 



in important buildings with corrugated galvanized iron and using 
galvanized-iron tie-roda, angles, and other construction shapes, in 
blind confidence of the protective power of the zinc coating; else in 
supreme indifference as to the future consequences and catastrophes 
that may arise from their failure. 



If 

ft 

n 



i 



Fio. 27.— Zincing solutiona recommended by viu-ious authorities. 

The comparative inertia of lead to the chemical action of many 
acids has led to the contention that it should form as good if not a 
better protection to iron than zinc, but in prnctice it is found to be 
deficient as a protective coating against corrosiofi. A piece of lead- 
coated iron or teme plate placed in water will show decided evi- 
dences of corrosion in twenty-four hours. This is to be attributed 
to the porous nature of the coating, whether it is applied by the hot 
or wet (acid) process. The lead does not bond to the plate as well 
as either of the other metals, zinc, tin, copper, or any alloys of them. 
Lead-coated iron corrodes rapidly when exposed to the gases of com- 
bustion. The usual weight of lead-coated teme-plates is about 
I ounce to a square foot, while hot-process zinc coatings weigh from 
li ounces minimum to 3 ounces maximum, depending upon the 
temperature of the bath, and the slowness of removal therefrom 
giving time for the article to drain off. The following table gives 
the increase in weight of different articles due to hot galvanizing: 



GALVANIZING. 



177 



Description of Article. 



Thin sheet iron - .026 inch No. 22 B. W. G. . 

Y*j-inch plates 

4-inch cut nails 

J-inch-dia. bolt and nut 



Weight of Zinc per 
Square Foot. 



1 . 196 oz. 

1.76 " 

2.19 " 
j approximately 
i 1.206 02. 



Percentage of 

Increase of 

Weight. 



18.2 
2.0 
6.72 

1.00 



Tin is often added to the hot bath for the purpose of obtaining a 
smoother surface and larger spangles or facets, but it is found to 
shorten the life of the coating considerably. 

A portion of a zinc coating applied by the hot process was found 
to be very brittle, breaking when attempts were made to bend it; 
the average thickness of the coating was .015 of an inch. 

An analysis gave the following result: 

Tin 2.20 

Iron 3.78 

Arsenic trace 

Zmc (by difference) 94.02 

A small quantity of iron is dissolved from all the articles placed 
in the molten-zinc bath, and a dross is formed amounting in many 
cases to 25 per cent of the whole amount of zinc used. This zinc-iron 
alloy is very brittle and contains by analysis 6 per cent of iron, and 
is used to cast small art ornaments from. 

Nickel coatings produced galvanicaUy will not protect iron from 
corrosion unless .02 inch thick. 

A hot galvanizing plant having a bath capacity of 10 feet by 
4 feet by 4 feet 6 inches outside dimensions, and about 1 inch in 
thickness, will cost $625, and will hold twenty-eight long tons of zinc, 
which at four cents per poimd will require $2500 to fill it; the heat- 
ing of this mass of metal and its ever-changing cold immersions, 
with the waste by dross and extra thickness in spots, is a constant 
source of annoyance and expense. 

The cost of an electro-chemical or wet-bath CJowper-Coles plant 
of 6700 gallons bath, size 30 feet by 6 feet by 7 feet, will be but 
slightly more than the hot bath given. There is no dross formed 
by the use of the Cowper-Coles process, and the zinc coating 
formed is said to resist the corroding action of a saturated solution 
of copper sulphate. The English Post-office test for telegraph wire 



178 



GALVANIZING. 



coated by the Cowper-Coles process shows much better than hot 
galvanized-iron wire, as per following table: 

Result of Process Test Made on Samples op Charcoal-iron Wire 

Coated with Zinc by Various Processes. 



Process Used to Coat the Wire. 


Grains of Zinc 

Per Square 

Foot. 


Ounces Per 
Square Foot. 


Number of One- 
minute Dips; 
Samples Stood 
without Showing 
Metallic Copper. 


Hot galvanized 


648.5 
446.4 
552.64 


1.48 
1.02 
1.26 


3 


Acid bath. ZnSO^ 


4 


Cowper-Coles process 


5 







A Cowper-Coles process bath of a capacity of about 4000 gal- 
lons will treat ship-plates 18 feet long, and will require an electrical 
energy of 2000 amperes of 5-volt electro-motive force. 

With equal amounts of zinc per unit of area, the zinc coating put 
on by the cold process is more resistant to the corroding action of a 
saturated solution of copper sulphate than is the case with steel 
coated by the ordinary hot galvanizing process; or, to put it in another 
form, articles coated by the cold process should have an equally long 
life under the same conditions of exposure that hot galvanized articles 
are exposed to, and with less zinc than would be necessary in the 
ordinary hot process. 

The hardness of a zinc surface is a matter of some importance. 
With this object in view, aluminium has been added from a separate 
crucible to the molten zinc at the moment of dipping the article to be 
zinced, so as to form a compound surface of zinco-aluminum, and to 
reduce the waste formed from the protective coverings of sal-ammo- 
niac, fat, glycerine, etc. The addition of the aluminum also reduces 
the thickness of the coating applied. 

Cold and hot galvanized plates appear to stand abrasion equally 
well. The thickness of the coating being the same, tests by means 
of the Schlerometer show: cold galvanized sheet, 6; hot galvanized 
sheet, 6; terne-plate, 2; tin-plate, 2. The figures represent the load 
in grammes upon a diamond point, just sufficient to cause it to scratch 
the specimen. 

The attempts to electro-zinc iron and steel wire for wire standing 
rigging, bridge, or other cables have not been successful; it has not 
been foimd practical to produce a wire capable of withstanding more 
than one immersion in a copper-sulphate solution. 



GALVANIZING. 



179 



Both pickling and hot galvanizing reduce the strength, distort 
and render brittle iron and steel wires of small sections. Zinc fuses 
at 775° F. and vaporizes at 830® F. Hence the necessity of the sal- 
ammoniac bath that covers the molten zinc, prevents volatilization 
and acts as a flux to unite the zinc and iron. The bath is usuaUy 
kept at about 1000® F. Steel wire of high breaking strain has its 

Table Giving Thickness of Zinc Required to Withstand Varying Num- 
ber OF Immersions in a Solution of Copper Sulphate. 



ut 
o 

z 

o 



^8 

Ik 
III 

< 

3 

3 

flC 

ut 

0. 



N 



Z 
O 

UJ 



— 2;«08 


9 
















/ 


•1 














/ 




1.974 

i tt4H 




Mr 












/ 






-1:010 












/ 








ttOf 


z 

< 

g 

A 








/ 










AfrA 


8 






A 












~.Go8 


2 


» 


/ 
















1 

i 


/ 




















J 


4 


2 : 


I 4 


i 


1 


i 


r 8 



No. of I miaate imaienioiiii in satuiatcd MMtttioa of oopper luIpliAlo 

Fio. 28. 

hardness, and consequently its ultimate tensile strength and elonga- 
tional efficiency, reduced by drawing the temper and the formation 
of an iron zinc alloy on the surface of the wire by as much as from 
5 to 10 per cent. It is the practice when coating steel wire to keep 
the bath at as low heat as possible and to run the wire through it 
at a high rate of speed. Both these operations lead to a waste of 
zinc by reason of the rapid solidification of the metal on the com- 
paratively cold wire, and consequently the ready breaking or crack- 
ing of the covering metal on bending or twisting it, owing to the 
difficulty with which molten zinc adheres to the steel except after 
long contact in the bath. In some cases the wire is wiped between 
asbestos rubbers as it leaves the bath, but wire thus treated is found 
to resist corrosion but a very short time. 



I 



180 OALVANIZING AND GALVANIC ACTIONS. 

English manufacturers have ceased galvanizing their high-grade 
steel wire that costs some $175 per ton, on account of the great risk 
of rendering it worthless. 

The Cowper-Coles or cold-galvanizing process is used for the 
purpose of zincing the skin plates and frames of the torpedo-boats 
and torpedo-boat destroyers built for the English navy. A plan 
and elevation of this plant is given in The Engineer, Feb. 28, 18M. 

The industrial importance of the successful application of this 
cold-galvanizing process can hardly be overestimated, even if its 
appUcation is only to the marine constructions of the future, and 
it is found to be in any degree inapplicable to our present structures 
and vessels in use. The permanency, continuity, strength, and den- 
sity of the coating given by this process is in all respects equal to 
that of hot galvanizing, and the thickness of it can be made superior 
to that given by the hot. Considering the success that has attended 
the use of zinc to prevent corrosion in marine boilers, where concen- 
trated hot saline fluids are the excitant medium, aided by the elec- 
trical conditions attendant upon the combustion of large quantities 
of fuel, it may not be considered a wild prophecy to expect that with 
all of the internal metallic parts of a steam vessel protected by an 
application of zinc plates secured to the framework of the structure 
similar to the application of zinc to marine boilers, that these i^lates 
may receive the energy of corrosion, and if not neutralizing it entirely, 
at least pass it along in the form of a deposit to convenient pockets, 
where it could be removed, the same as is now done with the wash- 
ings and dirt from the fire-room bunkers and ballast-chambers. 

This internal electro-chemical process of protection does not 
appear so chimerical as at first one might suppose. Dr. Henry Wurtz * 
has proposed the protection of mining plants subject to the intensi- 
fied corrosion due to the decomposition of pyrites and other minerals 
in the mine waters, by connecting all of the metal portions of the 
mine as the negative elements with a dynamo of suflicient force to 
overcome the strength of galvanic energy due to the surfaces exposed 
being excited by the corrosive liquids in the mine, the positive ter- 
minal to be connected to a mass of hard coke in the mine sump. 
These conditions vary but slightly from those existing in the ship, and 
it is not improbable that experiment will determine that both these 
S3rstems could be made to work successfully. 



* Engineering Magazine, May, 1894, Vol. VII, No. 3, page 297, "Preservation 
of Metals from Corrosion by Electric Polarization." 



ELECTRO-CHEMICAL AND GALVANIC ACTIONS. 181 

Thermo-electric currents arise from changes of temperatures in 
all bodies, and set up voltaic action in all cavities, fissures, seams, 
and contact surfaces in the metal, which, though slight and not 
easily detected, will in time enlarge and waste them away sufiiciently 
to sap the strength of the mass. 

Metallic salts and acids in mine waters intensify the corrosion of 
all metals exposed to their action. The metal work of railway tunnels 
is also disastrously affected by the condensed vapors of sulphur, 
carbonic acid, and the ever-present moisture due to such locations. 
The corrosion of the metals decreases the resistance of the water to 
voltaic circuits, this corrosion by liquids being voltaic phenomena in 
all cases. In many cases it is intensified by the moisture being in 
the form of drops instead of being uniformly spread over the whole 
surface. 

Acids and acid salts which are capable of taking up iron oxides 
into solution still further enhance the destruction by removing such 
oxides and exposing the surfaces of the metal to a fresh attack of 
the corrosive element. The saline matter in solution that excites 
voltaic action need not be acid. Any neiUral salt which decreases 
the resistance of the water will qualify it to act as the necessary liquid 
medium of a voltaic circuit. Sea-salt is the commonest of all such 
neutral salts, together with the other chlorides and sulphates of sea- 
water. It enables corroding voltaic action to be set up on all ferric 
bodies immersed therein or in the air impregnated with their substance. 

The Journal of the Society of Chemical Industry ^ London, February 
28, 1894, details some experiments with the galvanic action of sea- 
water upon iron and steel structures in various relations with each 
other, such as the constructive parts of trusses, boilers, etc., to pre- 
vent the corrosion for which the use of zinc and other easily oxidized 
metals and alloys are suggested, and to be so placed and connected 
to the structure that they will form the electro-positive element of 
the ever-present galvanic circuit, and by their decomposition protect 
the structure. 

Mr. D. Phillips, in a paper read before the Institute of Civil Engi- 
neers, in 1885, cited the result of an experiment, where "surfaces of 
bright pieces of plate iron, immersed in cold sea-water for over ten 
years have been thoroughly protected from corrosion by the aid of 
pieces of metallic zinc in metallic contact with the iron ; while a simi- 
lar piece of iron similarly fitted and immersed, but having a piece 
of paper placed between the iron and zinc plate, received no protection 



182 ELECTRO-CHEMICAL AND GALVANIC ACTIONS. 

whatever. The water was changed twice annually, and the oxide 
removed from the zinc by filing. Under these circumstances the iron 
became gradually coated with a film of leaden-colored deposit when 
wet, but hard and white when dry. The effect in other respects was 
that, on every occasion that the oxide was removed from the zinc and 
the deposit from the iron specimens, on being returned to the water 
small globules formed on the zinc, and on reaching -^ inch in diam- 
eter released themselves and flew to the surface." 

The proportions necessary to insure complete protection from 
corrosion in marine boilers are one square foot of zinc to fifty square 
feet of heating surface in new boilers, which may be diminished after 
a time to one in seventy-five or even one in one hundred square feet. 
Merely placing the zinc in trays, hangers, or strips will not insure 
metallic contact. The better and generally recognized method of 
fixing the zinc is to place a number of studs in the sides of the furnaces 
and combustion-chambers, and to bolt on to these studs the zinc 
plates, which should be about 10" X 6" XI". It is important to see 
that the contact surfaces are clean and bright, and the nut screwed 
close down to the zinc to exclude the water and deposits from the 
contact surfaces, thus comparatively insulating them and preventing 
the galvanic action. Otherwise the zinc is acted upon mostly as a 
solvent that renders the water innocuous or non-exciting, but does 
not prevent the water from forming a hard scale when it is saturated. 

Sheet zinc has proven to be a durable roofing material. Zinc 
is reduced in density from 6.86 to 7.2 in the process of rolling into 
sheets, which closes the pores and renders the metal less affected 
than tin-plate from the ammonia, carbonic acid and atmospheric 
gases. 

Berlin zinc roofing-plates (unpainted) have been foimd to be not 
materially affected after many years' exposure. The weather formed 
a thin film of oxide on their surfaces that effectually prevented further 
oxidation. A few cases reported are as follows. 

The Cloisters of Canterbury were covered with zinc roofing and 
were uninjured after 33 years. 

The Portsmouth Dock Yard Buildings' roofs were uninjured after 
24 years. The Great Western Railway Station roof at Rugby was 
uninjured after 20 years. Other railway-station roofs were uninjured 
after 15 to 20 years. The zinc roofing required painting if sulphurous 
acid was in the atmosphere. 

Tin roofing corrodes from the inside of the coating. It is also 



TIN, ZINC, AND LEAD ALLOYS, 183 

poroiis, quite as much so as a single coat of paint. The tin-plate as 
it leaves the molten dipping-bath becomes covered with a thin film 
of fluxing or non-drying oil that fills the pores of the tin, and if the 
roofing is painted soon as laid, this film prevents the paint from 
adhering to the tin, just as a machine grease prevents a paint from 
bonding to a surface. A few months' exposure to the atmosphere 
slightly oxidizes the tin, and this oxide absorbs the oily coating and 
allows the weather to wash it away and the paint has a clean 
metallic surface to bond to. Tin-plates doubly dipped are less por- 
ous and more diu'able. 

The quality of commercial tin-plate is greatly inferior to that made 
forty years ago, and appears to retrograde yearly. Lead, antimony, 
and other metals are mixed with the tin in the dipping-bath, and 
greatly reduce its resistance to corrosion. None of the adulterants 
form a true alloy; they are only mechanical mixtures. They all 
differ in oxidizing power and electrical affinities. The lead is electro- 
negative to the tin and zinc, which again are of opposite electrical 
natures. 

The amount of sulphurous and carbonic acids and ammonia in the 
atmosphere is enough to form the excitant element needed to decom- 
pose them one after another, until the coating is made porous 
and the iron is corroded in turn. The life of the tin-plates is also 
governed in a great measure by the want of care that they should 
receive in the preliminary pickling with muriatic acid, to free them 
from the mill-scale that always attends their rolling. Ordinary 
washing with lime-water does not remove the whole of this acid, and 
the tin coating usually has a double galvanic pile in a sandwich form, 
ready for duty on the least encouragement. 

There are brands of tin-plate as honestly coated at the present 
day, and as reliable in all respects, as any ever made, but they are 
an exception, not the rule. Price and the gullibility of the purchaser 
govern, as in many other modem industries. (See Fig. 6, page 38.) 

Red-lead paint coatings soften tin roofing, but do not wholly 
destroy it, although some of the tin may be changed to a white oxide 
that is easily removed by atmospheric influences. 



CHAPTER XVIII. 

INERT PIGMENTS, OR ADULTERANTS. 

The diflferent substances known as inert pigments are used to a 
great extent in the preparation of nearly all mixed paints, particularly 
in the house paints, where the amoxmt of one or more of them fre- 
quently exceeds that of the base pigment. 

However admissible their use (on accoimt of cost only) may be in 
paints not classed as protective ferric coatings, their durability in 
any case is determined by the character of the weakest element in 
the associated group to resist atmospheric conditions, whatever the 
base pigment may be. 

The manufacture of "patent pants'^ would be almost nil were it 
not for the very liberal use of these inert pigments. They are said 
to correct almost every detrimental quality in the basic pigments. 
Yet with all of their boasted virtues, there is hardly a manufacturer 
of paint willing to admit their use, or that will furnish an analysis of 
his product that contains them. 

Many of the uses and characteristics of these inerts, fortifiers^ 
or adulterants have been mentioned in the basic pigments chapters 
and elsewhere in this work, but are brought together here for com- 
parison and ready reference. 

Carbonate of lime (CaCo) in some form other than as quick- 
lime (calcined limestone or marble) is often used as a desirable adxil- 
terant of many paints. It is claimed to be specially favorable to 
correct the sulphur element present in iron oxides. 

Chalk is a friable carbonate of lime that, on accoimt of its cheap- 
ness and several colors, is the most used. Its specific gravity is 2.2 
to 2.8. According to the basic oxides in it, the colors are white, red. 
gray, and black. It contains about 2 per cent of clay besides free 
silica, magnesia, and chloride of calcium, and carries a large quantity 
of water. The latter is loosely held, and as calcination of chalk is 
not thought necessary when used for an adidterant, the moisture 

is carried into the paint to its detriment. 

184 



INERT PIGMENTS, OR ADULTERANTS. WHITING. 186 

Whiting. Spanish white and other trade-name whites are prepa- 
rations from chalk. When used to correct the sulphur element in 
iron-oxide paints, 10 per cent by weight of the pigment is often added. 
It is as easily dissolved by moisture as whitewash. Its use for the 
adulteration of white-lead pastes or paints is conunon rather than 
exceptional, and frequently composes 50 per cent of the paint. 
Whiting, when used as a pigment, is liable to form a chemical 
reaction between the oil and itself that results in the formation of a 
lime soap, which is not at all a durable substance. 

Putty, however (a mixtiu'e of whiting and oil), is a very durable 
body, and withstands atmospheric exposure and water remarkably 
well. In this form it illustrates the theory that the pigment is the 
life of the paint. The small amount of oil in the composition of putty 
is the cause of its quick drying. Its mass, when applied, greatly 
exceeds that of a paint coating, and its shrinkage solidifies, instead 
of rupturing it, by a movement in a number of directions, as in a 
paint. 

Barytes (Heavy Spar), Ba.S04. Specific gravity, 4.3 to 4.7. The 
natural sulphate of barium, consisting of one atom of barium oxide 
(BaO) = 65.67 per cent, and one atom of sulphuric acid, 34.33 per 
cent. It is the heaviest of all minerals, and is found in all stages of 
purity, in transparent, colorless, white to yellow crystals, also in a 
granular and compact form in heavy beds resembling marble. It 
is conunon in all metallic veins, allied with, or changed to, calc- 
spar, spathic iron ore, cerussite, quartz, limonite, pyrites, and other 
substances. It is the white variety that is ground for a pigment, 
but lacks the opacity or light-reflecting or coloring power to form of 
itself a good pigment. It grinds hard, splintery, and irregular, and 
is used to give weight to paper-stock, zinc oxide, gjrpsmn, and all the 
other light pigments that lack weight to enable them to masquerade 
as white lead. JiSee Chapter VI.) 

Barytes brightens light-colored paints, though of poor coloring 
or light-dispersing power; also spreads easily and saves oil. It is 
mixed with nearly all pigments, and by the use of a stiff or short 
bristle brush, covers a large siu^ace with a resemblance of a good 
paint. 

White lead does not cover so well with barytes, but zinc oxide 
covers better. Zinc oxide lacks weight that barytes furnishes and 
also saves oil, advantages not ignored by the cheap paint-com- 
pounders. Barytes does not unite with the oil in any degree. From 



186 INERT PIGMENTS, OR ADULTERANTS. BARYTES 

its weight and non-bonding nature, the paint is inclined to run on 
vertical surfaces. This tendency requires the use of large amounts 
of volatiles or quick driers. Barytes alone is the poorest of pigments. 

Floated barytes is the ground natural sulphate of baryta floated 
off in water to give a finer product. 

Artificial barytes (Blanc-Fixes) is made by heating barium car- 
bonate with sulphuric acid and precipitating the artificial barytes 
from the solution. It is less crystalline than the natural sulphate 
and has a greater covering power. 

Baryta white, permanent white, constant white, etc., are of 
this class of pigments. Blanc-Fixe mixed with the mineral barytes 
compose the principal substances in most of the commercial white 
patent paints. 

Lithopone, a trade-mark for one of these mixtures, has an extended 
sale under the guise of white lead. Trade-marks are easily invented, 
but they add no durability to the paint. They move around in paint 
literature as easily as some of the substances covered by the name 
move in the vehicle that gives them a home, if not rest. 

Barytes as a pigment, exposed to air or on underground bodies, 
condenses water and carbonic acid and is converted into a carbonate 
with the evolution of sulphuretted hydrogen. This decomposing 
feature in barytes seems to be ignored by paint-compounders, but 
to it the failure of many coatings can be attributed. 

Rose's experiments show that barytes in any form, when acted 
upon by water, evolves sulphuretted hydrogen and sulphurous acid, 
leaving the decomposed lime free. 

Hansfeld has also shown that sulphate of lime is decomposed by 
the galvanic action of two metals or metallic oxides in contact under 
the ordinary exposures of a paint. When both barytes and gypsum 
are present in a paint, this galvanic action between three substances 
is certain to occur. Mixtures of barytes and gypsum with the oxides 
or carbonates of zinc or lead will in no degree protect any one of 
them. They decompose one after the other; the first to break down 
only adds to the electrolytic energy to hurry up the decomposition 
of the others. The thin film of oil in which the pigments are embedded 
is sufficiently porous to admit the atmospheric moisture and carbonic 
acid necessary to start up the disintegrating process. 

Putty made from barytes, whiting, and oil dries into a hard, 
brittle mass that crumbles easily. Glycerine added to it only tem- 
porarily defers the extreme shrinkage and crumbling. 



ISERT PIGMESTS, OR ADULTERANTS. BRICK-DUST. 187 

Probably the greater number of the mixed white-lead pastes and 
paints sold in the world contain barytes; as it costs only 110 to $20 
per ton, or about one-sixth as much as white lead or zinc oxide, the 
temptation to use it in place of these pigments ia not always resisted. 
As a rule no responsible paint firm with a business reputation to 
sustain will sell a barytes adulterated paint under their name. 

Additional data about the presence of barytes in a paint is given 
in Chapter XXIX. The covering and coloring power of barytes in 
comparison with white lead and zinc oxide are shown by Fig. 29. 



Fig. 29. — Covering power of incrl pigments. 

Brick-dust is used to a large extent to adulterate red lead and other 
red paints ; even the low-price iron oxides do not escape it, particularly 
the brighter-colored copperas oxide. It tones them down to many 
required trade-shades. Care is not always exercised to grind only 



188 INERT PIGMENTS, OR ADULTERANTS. FELDSPAR, 

the hard-burned bricks for the pigment, hence many of the samples 
are not much better than a dried clay or red chalk. 

Prof. Mallett's experiments with paints composed of pulverized 
hard-burned red tiles, iron oxide, and red lead were favorable in cer- 
tain proportions of the several substances, and decidedly unfavorable 
with other proportions of the same ingredients, whether applied 
to wood or iron. In general the tiles added nothing to the quality of 
the paint, only reduced the cost of it. They are practically unoxidiz- 
able by atmospheric influences or weak acidulous solutions, and are 
electronegative to metals or their oxides. In any electrolytic action 
set up by any cause in a paint coating of which brick-dust is a part, 
the tendency will be to decompose the other pigments, possibly, before 
electrolysis is developed in the covered iron surface. (See page 53.) 

Feldspar. Specific gravity, 2.5 to 2.8. Decomposed mica, granite, 
gneiss, and most forms of basalt form this class of adulterants, and 
are all inclined to further decomposition on exposure to the weather. 
Many of the fire-clays used in the manufacture of fire-brick are broken 
down and decomposed feldspars. Its use in the composition of a 
pigment is of the most unreliable character in all respects. It is as 
poor a substance for an adulterant as nature furnishes. Mixed paints 
frequently contain 15 to 20 per cent of it. Feldspar carries a large 
amount of water loosely held and frequently acidulated, also sand, 
etc. It is easily whipped up in the oil and mixes well with graphite for 
dark-colored paints. 

Gypsum (Sulphate of Lime). Natural mineral (hydrated), CaO.SO., 
+H3O; calcined, CaO.SOj. Specific gravity, 2.4 to 2.8. The nat- 
ural mineral ground is the plaster that farmers use on their crops 
to attract and condense atmospheric moisture. Calcined to expel 
the one atom of water held in its natural state, it becomes the 
common plaster of Paris, used for the hard finish of plastered walls 
of buildings. The process of grinding is supposed to drive off the 
one atom of water it holds natiu-ally, by the heat developed in the 
dry grinding-mill, but this is soon replaced upon a short exposure 
to the atmosphere, and when used for the hard finish, it must be 
heated again to dispel it, or a porous wall coating results. A hydrated 
sulphate of lime contains over 18 per cent of water. 

When used in the composition of a paint, it must be thoroughly 
calcined, and is so specified by parties who allow its use. Neglect 
of this carries the moistiu^e into the paint, where some portion of the 
sulphur element in the gjrpsum is released, and combining with the 



l^^ERT PIGMENTS, OR ADULTERANTS. GYPSUM. 189 

fatty acids in the oil, sometimes causes the paint "to liver," a phe- 
nomenon famUiar to all painters, but not always attributed to the 
right cause, viz.: too much of a sulphurous adulterant. 

Gypsum grinds easily, is opaque, and incorporates readUy with 
most pigments and the vehicle. It is not liable to set or settle in the 
paint-bucket or package, and probably is the best of all the inert 
substances to use as an adulterant. 

The extra atom of sulphur in the natural mineral other than that 
necessary to form the sulphurous-acid compound is strongly com- 
bined, but if the mineral or ground pigment is calcined at a tempera- 
ture higher than that necessary to release the one atom of water, 
the sulphuric-acid atoms are excited to a degree that will afterward 
manifest the same destructive properties as the same element does 
in any other pigment or substance, and as noted above in the '*liver- 
ing" of the green-paint coating. 

A synthetical sulphate of lime is supposed to be formed when 
iron ore is roasted in a furnace in contact with a quantity of carbon- 
ate of lime. The roasting process, besides driving off the moisture 
in both the iron ore and carbonate of lime, and a part of the sulphur 
in the iron ore, excites the remaining atoms of sulphur to leave the 
ore, and combines ^dth the now anhydrous carbonate of lime, CaO, 
and forms the anhydrous sulphate of lime, CaO.SOj, described above. 
The process is an unsatisfactory one, as the carbonate is generally 
added in great excess of the amount needed to effect the chemical 
combination with the sulphur to form the sulphate. When the roasted 
iron ore is removed from the furnace to be ground, the sulphate is 
not distinguishable or separable from the uncombined carbonate of 
lime, and both are ground with the ore and appear as adulterants, 
that may be 5, 10, or 20 per cent, or as much as can be unloaded 
upon the consumer. 

In whatever amount these lime products are present in the iron- 
oxide pigment, they are both, like the oxide, anhydrous, hygroscopic, 
and readUy attract moisture, frequently 5 per cent in oxide pig- 
ments that have been made for some time. 

The synthetical sulphate of lime formed in the roasting of cop- 
peras, as described in the preparation of that substance for an iron- 
oxide pigment, is of the same character as that from roasted iron 
ore, only it carries more loosely combined sulphuric acid in its com- 
position, as is denoted by the brighter color of the pigment. Its 



190 INERT PIGMENTS, OR ADULTERANTS. KAOLIN. 

effect upon the paint coating is the same, and is not conducive to 
any permanency of color or protective qualities. 

Kaolin {American Terra Alba). A clay of the same class as pipe- 
clay, China clay, potters' clay, etc. The reddish color of the latter 
being due to the iron and other metallic oxides. Specific gravities, 
2.58 to 2.76. 

Their general composition is as follows: 



Substances. 



Alumina 

Silica 

Lime 

Magnesia 

Oxide of iron. . . 
Alkaline earths. . 
Sulphate of lime. 
Moisture 



Pipe-clay, 
White. 



Percenta^ses. 
23.2.') to 21.28 
72.23 " 65.49 



12.00to 4.00 



Potters' and 

China Clay, 

Light and Dark. 



Percentages. 
23.25 to 21.28 
73.33 " 64.95 



1.26to 2.54 
7.26 " 1.78 
4.72 " traces 
4.00 " 1.30 



Terra-Alba, 
White and Gray. 



Percentages. 

28.51 to 15.50 

67.50 " 49.65 

22.40" 1.52 

0.17 " traces 

3.70 " 2.05 



8.58to 0.42 



All of the pigment-clays grind greasy, and are as easily broken 
down and decomposed by the weather as any of the clays in building 
brick or mud from a mill-pond. 

Mixed with talc, the clays are supposed to add some advantage 
to pigments of a granular character. What that advantage is the 
author has never been able to ascertain, but he knows they cause 
paint to peel or crack. 

Marl. Specific gravity, 2.4. It is composed of: 

Carbonate of lime 50 per cent. 

Silica 12 " " 

Alumina 32 " " 

Oxides of iron and manganese. . . . 2 " " 
Water 4 " " 

100 " " 

Its gray color prevents its use as an adulterant of the white paints, 
but in the tinted colors it is used quite as freely as kaolin or chalk. 
It is difficult to pulverize on account of its greasy nature. It saves 
oil, but causes the coating to chalk or peel on a short exposure to 
atmospheric influences. 



INERT PIGMENTS, OR ADULTERANTS. OCHRE. 191 

Ochre. A yellow clay containing from 8 to 15 per cent of water 
loosely held with laige amounts of sand, also marked quantities of 
iron oxide and sulphur. When moderately heated, the lower grades 
of ochre contain sulphur enough to change their color to a pink or 
low red. 

The common grades were formerly used as a coating for tin roofs. 
They were always subject to blistering, from the laige quantity of 
water they carried into the paint. It is an adulterant without a 
single inert element in it, and its presence in a paint is generally 
accompanied by as poor a quality of oil as it is a pigment. 

SUica (Si.SOj). Specific gravity, 1.9, 2.5, 2.8. Is a sulphate of 
silicon containing one atom of silicon combined with one atom of 
sulphurous acid. It is found in crystals of different degrees of 
translucency, and forms a component part of all metallic ores; 
of iron ore, frequently, 50 per cent. (See Analyses of Iron Ores, 
Chapter III.) 

In its natural form it is one of the most imperishable of all min- 
erals. It grinds hard and splintery, and is difficult to reduce to the 
fineness required for a pigment. Manufacturers of silica products 
subject the crystals to a bright-red heat and quench them in water, 
causing them to fracture and grind more easily. However, the 
calcination drives ofif a part of the sulphuric acid and renders the 
silica caustic, — the latter condition is not a favorable one for any 
pigment, inert or basic. 

Silica is not affected by sulphurous gases, acids, or alkalies. 
Floated silica or silex makes an excellent wood-filler paint. All 
silicas are difficult to hold up in oil, and on settling, cake very hard. 

Sand, generally supposed to be the same substance as silica, is, 
however, of quartz formation as an oxide of silicon, specific gravity 
1.44 to 1.76, only about two-thirds the weight of silica. It grinds 
hard and splintery, or in an irregular crjrstaDine form, and is difficult 
to grind to a pigment. 

Neither silica nor sand mix with other pigments, except in a purely 
mechanical manner, differing according to the specific gravities of 
the several substances incorporated together. They have no affinity 
for the vehicle, are not in their pulverized form absorbent of moisture, 
except to a very small amount. From their indestructible nature, 
they form the electro-negative element or centres to determine the 
electrol)rtic action always present in the decay of a paint coating. 
Their use as adulterants of mixed paints is greater than any manufac- 



192 INERT PIGMENTS, AMOUNTS MANUFACTURED, ETC. 

turer of paints will acknowledge. The covering power of silica is 
shown by Fig. 29. 

Fine sand is used for an application to green paint on exterior 
surfaces with a view to affording an extra protection from atmos- 
pheric effects. It in all cases hastens the decay of the paint. It 
holds the moisture, dust, and other organic substances, and their 
easy and early decomposition results. Blisters form more readily 
imder a sand-coated paint than with a paint alone. 

Talc (Steatite or Soapstone), Specific gravity, 2.65 to 2.8. 
Grinds greasy and flaky, is inclined to cause a paint to peel, and is 
repellent to the oil. Its use with kaolin has been given. It is used 
for a special adulterant of flake and other graphitic carbons (see 
Chapter XIII). As a pigment it has no qualities whatever. As an 
adulterant its function is to enable some objectionable substance to 
attempt a mission that could be better performed by a straight pig- 
ment. 

The above list does not exhaust the substances known as adul- 
terants or miscalled " inert ^* pigments. As protective coverings for 
ferric bodies, the protective effects of the inert pigments in use with 
the basic pigments will be noted in the paint tests made to determine 
their value (see Chapters XXIX and XXX). Their covering or 
coloring powers are shown by Fig. 29. 

The following amounts of inert pigments were produced and used 
in the United States (averaged for the years 1898 to 1901) : 

Barytes, all grades, 124,000 short tons. Cost of the crude mineral, 
S3.30 to $3.50 per ton. 

Imported barytes, 1400 tons, including some floated. Cost of 
the manufactured article, $10.50 to $11.00 per ton. 

Feldspar mined in the United States for all purposes in 1898 to 
1901 averaged 27,280 tons and cost from $3 to $6 per ton. 

Ground slate and shale for pigments averaged 4,700 short tons, 
value from $9.50 to $10.00 per ton. 

Of soapstone ground for pigments and foundry use, there were 
produced 9000 tons yearly. Value 8^ to 9 cents per pound. 

Of gypsum calcined, for all purposes, 280,000 short tons were pro- 
duced in 1901, cost $3.50 to $3.90 per ton. 

Crude gypsum costs $1.20 to $1.25 per short ton. 



CHAPTER XIX. 

SPIRITS OF TURPENTINE. 

The composition of spirits or oil of turpentine is C,oHie. Specific 
gravity, 0.86 to 0.88, with a boiling-point always near 160® F. The 
several varieties of commercial turpentine obtained from the sap of 
fir- and pine-trees are more or Jess viscid solutions of resins in a vola- 
tile oil, the proportions of these constituents varying according to 
the source and age of the turpentine-tree. Some kinds are clear and 
homogeneous; others are more or less turbid, holding in suspension 
granulo-crystalline masses, which gradually settle to the bottom, and 
are known to painters as ** drops." 

Spirits of turpentine is the product of the first distillation of the 
crude gum, and consists of about one-third spirits and two-thirds water. 
It requires about twenty-five barrels of crude gum to make two 
barrels of the spirits of turpentine, that when redistilled is known 
as refined or oil of turpentine. 

The principal supply of turpentine is obtained from the American 
long-leaf yellow pine-tree, Pinus pcdustris (P. australis); also from 
the loblolly-pine, P. tceda; all products of the southern part of the 
United States, where the Coniferae are the principal trees. There are 
many varieties of the Coniferae, and all yield gums available for distil- 
lation into turpentines and resins. 

Turpentine consists chiefly of a hydrocarbon oil (CioHje) and a 
resin called "Colophony" (C20H4PO3). Specific gravity, 1.07 to 1.08. 
It softens at Ibb"" to 175** F. and melts between 194** and 212** F. 
The spirits of turpentine constitutes about 17 per cent of the yellow 
pine-tree sap or crude gum. The Maritime pine furnishes about 24 
per cent of spirits of turpentine. 

The exuded gum from all of the turpentine-trees is a yellowish, 

opaque, tough mass, brittle and crumbly when cold, crystalline in 

the interior, and of a characteristic taste and odor, a distinguishing 

feature in all types of the " turpens," designated as the terebinthic 

odor. The commercial oils of turpentine are as follows: 

193 



194 SPIRITS OF TURPENTINE 

The German, derived chiefly from the Pinus aylvestris (Scotch fir> 
P. nigra, and P. rotundata. 

The English, from the American or Carolina Pinus auatralis or 
P. tceda. 

The French, or Bordeaux, from the Pinus maritima, resembles 
the American turpentine in appearance, odor, and taste, and is con- 
sidered to be the quickest drier. 

The Strasburg is the product from the Abies pectinata and from 
the spruce fir, Abies excelsa. 

The Venice is the product from the Terebinthina venita, or the 
larch, Larix europea. 

The Hungarian is from Pinus pumUio. 

The Carpathian, from the Pinus cefmbra, has a bitter taste. 

The Cyprian, Syrian, or Chio, obtained in Chio, is from the Pistada 
terebinthus, 

Templin, or pine-cone oil, is furnished from the cones of the Pinus 
pumUio and the Abies pectinata. 

The Canadian oil or Canada balsam, from the Abies balsamea (Balm 
of Gilead), furnishes the whitest and purest of all of the turpentines. 

Related to the true turpentine-oils are the two volatile oils of the 
coniferous plants — oil of juniper from the Juniperus communis, and 
the oil of savin from the Juniperus sabina, 

A characteristic feature in American turpentines is that they 
polarized to the right, while most of the turpentines from other 
sources polarize to the left. 

The crude resin from which the oil of turpentine is distilled has a 
specific gravity of 0.95 to 0.98, according to the time of its collec- 
tion, whether in the first, second, third, or fourth year after the tree 
is boxed, or during the time of collecting the dried sap in the first 
flow in the spring, or the summer, or later in the fall. Also its freedom 
from sand, leaves, bark, and dirt; all of which are readily absorbed 
by the sticky, drying sap, and are only removed in the process of 
distilling the gum for spirits of turpentine. Another distillation is 
required to produce the oil of turpentine or "turps" of the painter. 

Fig. 30 shows the old method of boxing the trees to collect the 
crude resin. 

In addition to the exuded gum from living trees, turpentine is 
also obtained by the distillation of the dead wood from the long-leaf 
pine-tree when it no longer yields the gum (after the fourth or fifth 
year), and when it has been turned over to the lumber or cord-wood 



SPIRITS OF TURPENTINE. 195 

men. The trees not fit for lumber, or unfavorably located for hand- 
ling the cord-wood, are cut down and distilled in a kiln or oven similar 
to that used for the production of charcoal from hard wood. 



Fig. 30. — Boxing the turpentine-tree. 

A cord of fat pine-wood yields by kiln distillation, according to 
the amount of the pitchy matter in it, whether it is body wood or 
from the limbs, tops, or decayed wood, the following products: 

Turpentine crude oil 22 to 26 gallons i 

P\'ralifi;neous acid 86 " 90 " [â–  1 150 to 1200 pounds. 

Kne-tar 118 " 122 " ) 

Charcoal 54 " 68 bushels— 2200 to 2400 pounds. 



196 



SPIRITS OF TURPENTINE. 



Many commercial turpentines contain acid. They are generally 
the products of kilns, and are not redistilled to free them from the 
acids. The effect of the use of turpentine in a paint or varnish is 
to flatten the gloss or lustre. 

Even with a pure turpentine not more than 3i per cent is admissi- 
ble in a paint; and less than this if the turpentine is poor or fatty. 

Pure turpentine-oil is adulterated with crude or undistilled tur- 
pentine, light-colored resin-oil, and resin. 

These adulterations are detected by the pyroligneous smell and 
nauseous after-taste on the tongue and by the change in the specific 
gravity. Also by the reaction produced by adding 8 drops of strong 
ammonia to 90 c.c. (1.422 cubic inch) to the turpentine. The follow- 
ing are the results: 



Pure Oil of Turpentme. 



Specific 
Gravity. 



Reactions with Ammonia. 



Pure oil of turpentine recently 

distilled 

7 .4409 pounds per gal. 



Old pure oil of turpentine 

7 . 4534 pounds per gal. 



Pure turpentine with 10 per 

cent of resin spirit 

7 . 3496 pounds per gal. 

Pure turpentine with 10 per 
cent of undistilled turpen- 
tine 

7 . 3496 ik)unds per gal. 

Pure turpentine with 10 per 

cent of repin 

7.3686 ijounds per gal. 



0.8678 



0.8693 



0.8784 



0.8784 



0.8831 



No effect. The turpentine evaporates 
quickly. No residuum. 



Solidifies in a few seconds, forming a 
white crystalline substance with the 
consistency of butter. 



Forms an emulsion, which rapidly be- 
comes clear. Tne anmionia which 
separates has a ]>ale-yellow color. 



Forms an emulsion which becomes 
clear on standing, gives a semi-trans- 
parent sediment of a bluish color, the 
liquid above being colorless. 



Each drop of anunonia appears to solid- 
ify as it falls into the oil. On agita- 
tion the whole solidifies into a consis- 
tent transparent mass. 



Characteristics of Oil of Turpentine. 

Pure oil of turpentine has the composition of Qo Hn, and at a 
specific gravity of 0.839 weighs 7 pounds per gallon. At 60T. 
the gravity is 31° Baum^, and it weighs 7 pounds per gallon. At 
60° F. pure turpentine should weigh not less than 6.802 nor more 
than 7.278 pounds per gallon. 

Benzine at 60° F. has a gravity of 65° to 72° Baum6 and weighs 
6^ to 6 pounds per gallon. The hj^drometer test for benzine is 62° 



SPIRITS OF TURPENTINE, 197 

BaumS. Any benzine or light kerosene-oil added to turpentine will 
raise the degree to some point between 32® and 65® B. 36® to 38® B. 
should be the limit of acceptance for turpentine. 

Turpentine adulterated with mineral oil will leave a stain on a 
blotting-paper filter. 

An average quality of turpentine boils at 320® to 350® F. and 
has a flash-point of 103® or 104® F. 

Crude turpentine resin boils at 316® F. 

Crude turpentine resin, specific gravity 0.98 to 0.95, is dissoluble 
in water, but readily soluble in ether or spirits of turpentine and 
in six parts of alcohol. The alcoholic solution has an acid reaction. 

Bromine and iodine act violently upon it. When brought into 
contact with a mixture of nitric and sulphuric acids it takes Are. 
Turpentine is a solvent of all oils and resinous gums at ordinary tem- 
peratures, but some of the fossil resins require a low heat to aid its 
action. 

AdvUerants of Turpentine. 

Kiln-distilled spirits of turpentine contains pyroligneous and 
other acids, specific gravity, 0.80 to 0.84. 

Crude petroleum, specific gravity, 38® to 48® Baum6; weight, 7.00 to 
6.62 pounds per gallon. 

Benzine, specific gravity, 54® to 62® B.; weight, 6.39 to 6.10 
pounds per gallon. 

Naphtha, specific gravity, 62® to 70® B. ; weight, 6.09 to 5.79 pounds 
per gallon. 

The pyroligneous acid in turpentine distilled from dead-fat pine- 
wood settles out partiaDy after standing, but commercial brands not 
redistilled still contain some amount of the acid. 

Commercial spirits of turpentine has a specific gravity of 32® 
Baum6. Any addition of petroleum of 40® B. will be shown by the 
rise in the hydrometer; each 3 to 5 per cent of petroleum added 
causes a rise of 1® B. on the scale. If the adulterant is benzine or 
naphtha, then the difference in the specific gravity is very marked and 
will at once determine the character of the adulterant, it being much 
lighter than coal-oil or kerosene. 

For the detection of resin in the spirits of turpentine, the polari- 
scope-test is the only one that can be considered strictly accurate, 
but it is a delicate one, and requires experience to determine results. 

A ready method for detecting resin consists in dilute sulphurio 



198 SPIRITS OF TURPENTINE. 

acid one part and four parts spirits of turpentine; mix and shake in 
a test-tube and notice the precipitate, which will be the resin; allow 
for the resin normal in all turpentines; the pure spirits of turpentine 
will be found on top of the fluid in the tube. The flash-test for naphtha 
adulterations consists in heating the turpentine in a double vessel 
63° to 65° F. and then flashing it. If it ignites, it is safe to assxmie the 
mixture contains more or less naphtha. 

. The Journal of Chemical Industry , Vol. IX, 1890, p. 657, gives a 
test for commercial spirits of turpentine, the usual adulterants being 
naphtha, petrolemn, resin-oil, and the inferior Russian oil of turpen- 
tine. (See also same Journal, pp. 330-557.) 

The U. S. Navy Department tests for turpentine are: A single 
drop placed on white paper should completely evaporate at a tem- 
perature of 70° F. without leaving a stain. 

A few drops on a piece of white paper, hung vertically before the 
light, if the turpentine is pure and well distilled, should leave no mark 
after 5 to 7 minutes. A faint mark indicates the presence of resin 
due to imperfect distillation. If a gray mark remains for an hour or 
more, it indicates kerosene or other petroleum oil. If a greasy mark 
remains over 10 to 12 hours, petroleum is present in large quantities. 

It requires from 680,000 to 1,000,000 acres or 1060 to 1560 square 
miles of forest to supply the turpentine products, whose value is from 
$8,000,000 to $8,600,000 yearly. Both the export and home demand 
are increasing from 5 to 8 per cent yearly, and the forest supply 
for tapping is decreasing in more than an arithmetical ratio of these 
amounts. 

The unit of product for a turpentine crop is 10,000 boxes 
of 2500 trees from 100 to 300 acres of forest, according to the 
size of the trees, or an average of 15 trees per acre. When the 
lumber is exhausted and the cord-wood is cut out, there remains 
about one-half a cord of wood per tree available for kiln distillation. 
This will 3deld about 12 gallons of crude turpentine spirits that would 
redistill to about 10 gallons of the oil of turpentine per tree. This 
amount, if it all could be collected and distilled, would yield about 
6 times the yearly demand of 22,000,000 gallons. But 95 per cent 
of this supply of wood would be used for saw-mill fuel, cord-wood, 
waste, and be unavailable on account of location, or standing as dead- 
wood forest. The latter is a fruitful source of the forest fires that 
annually destroy from 3000 to 5000 acres of this valuable timber. 
Hence but 5 per cent would find its way to the kiln, and fumidi 



SPIRITS OF TURPENTINE. 199 

about 25,000,000 gallons of turpentine, or a little more than the 
present (1903) requirement, if the supply came from this source 
alone. 

When the long-leaf pine forests have practically disappeared, 
they will have to be carefully gleaned once and for all in order to 
produce a quantity of turpentine equal to the present demand for 
one year. 

A barrel (240 to 260 poimds) of the crude turpentine resin, when 
distilled, yields from 10 to 11 gallons of turpentine spirits that need 
to be redistilled to afford a pure oil of turpentine. About one-half 
or five-eighths of a barrel of resin (170 to 190 pounds) is also a re- 
sult of this distillation. This resin is redistilled for resin-oils of a 
number of grades, whose specific gravities range from 0.960 to 0.9910. 
It also furnishes sixteen recognized grades of conunercial resins; those 
known as W. W. (water-white), W. G. (window glass), are the finest 
and most valuable, being produced from the first year's run or virgin 
sap. Each subsequent year of the four or five years that the trees run 
resin, an inferior quality is produced, that is graded N. (very clear) ; 
then M. L. K., J. H. to A., the latter being almost black, and rated 
conmiercially as pitch, specific gravity, 1.15. 

The flow of resin from the freshly boxed or virgin cut trees is from 
250 to 350 barrels of 240 to 260 pounds for the first year, and requires 
100 to 200 acres of forest; the flow decreasing to 48 or 60 barrels in 
the fourth year, that furnishes the poorer grade of crude resin, that 
contains but little turpentine. 

The action of turpentine as a drier for paint or varnish is to form 
the peroxide of hydrogen from the air that renders them non-drying 
except upon the surface. 

Turpentine, by absorbing oxygen from the air as it stands in the 
barrel, is liable to become "fatty" (Cio.Hjfl.Oj) with age, and cannot be 
properly corrected except by redistilling. The use of such turpentine 
in a paint is to render it "tacky." Painters resort to the use of ben- 
zine to correct this fatty condition, but it is detrimental to the life of 
the paint and to its gloss. 

Fatty turpentine evaporates slowly on blotting-paper and leaves 
a stain upon it. 

A drop of turpentine allowed to spread itself slowly down a piece 
of glass coated black upon the other side of the plate will show a bluish 
tinge at the edges if petroleirai is present even to the amount of 6 per 
cent. 



200 SPIRITS OF TURPENTINE. 

Adulterations of turpentine with resin-oil are shown where the 
residue left after evaporating a small quantity in a saucer is of a 
sticky nature and resinous odor after it is ignited. 

The use of tank-cars that have been used to transport crude petro- 
leum is responsible for a great deal of the impure oil of turpentine. 
The crude oil of turpentine thus transported and carelessly redistihed 
will carry over enough of the petroleum to sensibly raise the specific 
gravity of the turpentine. 

From a large number of tests of commercial turpentines by various 
State associations of painters, and by individual painters and experi- 
menters, the general result appears to be that 50 per cent of the 
samples showed adulterations ranging from 5 to 20 per cent. There 
are no penalties for the adulteration of either turpentine- or linseed- 
oil, and when adulterations are present in any form or amount, gen- 
erally, the detection of them is beyond the power of the ordinary 
purchasing agent or painter. 

The Secretary of Agriculture for the United States for the year 
1890 reports that at the present rate of consumption, the forests of 
the long-leaf or turpentine pine will be exhausted in from 8 to 10 
years. Practically the yellow-pine forests of North and South Carolina 
are exhausted, and the production of turpentine and resin is now 
centred in Georgia, Alabama, and Florida. The belt of long-leaf 
pine timber extends about 150 miles inland from the seacoast across 
the above States to the Mississippi River. Texas has been com- 
paratively denuded of the yellow pine. 

Where the supply of turpentine and resin will come from when 
these forests are extinct, is an unsolved problem. The second growth 
of timber following the pine appears to tend toward the scrub-oaks 
and non-resinous trees — cedars, etc. 

Fig. 31 shows the modem or improved method of scarfing the tur- 
pentine-tree and collecting the sap, as distinguished from boxing 
the tree. 

Over 75 per cent of the turpentine produced in the United 
States is exported. Europe has no yellow-leaf pine forests that 
furnish any great amount of turpentine. Norway, Sweden, and 
Russia furnish resins from the other pine varieties of the Conifei«, 
but they rate low in the amount of turpentine they contain as com- 
pared with the American or hot-belt growth of the long-leaf yellow 
pine. 

The exports of spirits of turpentine of all grades from the United 



SPIRITS OF TURPENTINE. 201 

States for the year 1897, to only six of the European countries, were 
as follon's: 

Austria 65,000 gaUona. 

Belgium 2,088,810 " 

Germany 2,418,790 " 

Italy 398,710 " 

Netherianda 2,359,590 " 

Great BriUin 8,476,700 " 

Total 15,817,600 " 

Other countries 682,400 " 

16,500,000 " 
Uiut«d States consumption 5,500,000 " 



r 



22,000,000 " of turpentine, and 

Total ptoduction. -J 1,600,000 barrels of resin— 

350 to 400 pounds per barrel 



Fig. 31. — Boxing the turpentine-ttee. New method. 

Mr. F. G. Frankorter* reports "that the products of the pitch 
made from the butt of the Douglas fir or Or^on pine are unusually 



" Science," July 24, 1903. Avieriean Chemiad SoeUtg Journal. 1903 



202 SPIRITS OF TURPENTINE. 

rich in pitch. They contain as high as 41.6 per cent, of which 
21 per cent is turpentine. The latter has about the same boiling- 
point (150'' F.) as that from the northern pine, but differs from it in 
other properties. The kiln products (turpentine, pyroligneous acid, 
charcoal, pitch, etc.) from one butt discarded as imfit for liunber 
had a value of $275." The tree grows on any mountainous soil 
where the spruce, hemlock, or any Coniferse grow, and may be here- 
after utilized for its turpentine-supply only. It grows from British 
Columbia to Mexico in large forests, the trees often reaching 300 feet 
in height. The bark is useful for tanning. 



CHAPTER XX. 

CARBON BISULPHIDE 
(carbon DISULPHIDE — SULPHO-CARBONIC AdDS). 

This carbonic anhydride (CS,) has a specific gravity at 32® F. of 
1.027 to 1.072. At 60** F., 1.272 or 10.6136 pounds per gallon. The 
specific gravity of its vapor at 60^ F. is 2.6292 to 2.644. The 
boiling-point of the commercial article is 118.4*^ F., that of the refined 

109.4® F. 

It is composed of 15.8 per cent of carbon and 84.2 per cent of sulphur 
(CSj), and is produced by passing the vapor of burning sulphur (sul- 
phurous-acid gas, SOj) over charcoal kept at a red heat. This is the 
commercial method of manufacture. It is highly inflamma ble; its 
vapor mixed with air takes fire at about 300° F. and explodes with 
great violence. It is a colorless, heavy, very volatile liquid, possessing 
an acid, pungent taste and a very fetid alliaceous odor, due to the 
impurities of 8 to 10 per cent of sulphur and hydrogen compounds in 
the unrefined product. When refined, it loses the nauseous smell 
And has an ether-Uke odor, but it is necessary to keep it under water 
in air-tight iron vessels. 

Carbon and sulphur do not combine when simply heated together 
in the solid state, because the sulphur volatilizes before the necessary 
heat is attained. But when the charcoal is ignited to redness and the 
sulphur vapor is passed over it, CS2 is formed. 

Carbon bisulphide is deadly poisonous; inhalation of even very 
•dilute vapor producing giddiness and vomiting, with irresistible fits of 
weeping, violent pains in the legs, and a collapse of all the bodily 
and mental powers; paralysis, idiocy, and death. Five per cent of 
the vapor in any confined space ensures the death of all larvae, smaller 
mammalia, birds, and reptiles. A solution of ferro-carbonate in car- 
bonic-acid water is a partial remedy for the symptoms on first attack. 

It is a solvent of all fats, oils, resins, india-rubber, phosphorus, 

bromine, chlorine, iodine, camphor, etc., and mixes in almost all 

203 



204 BISULPHIDE OF CARBON. 

proportions with alcohol, ether, benzine, and all the fixed and volatile 
oils. 

It is used to extract the oil from seeds, particularly linseed, 
which is heated and pressed to remove some of the oil before being 
submitted to the action of the bisulphide. The residue cake contains 
only 2 per cent of oil and about 7 per cent of water, while the cake 
from the ordinary process of manufacturing linseed-oil from the 
steamed linseed contains 9 per cent of oil and nearly 15 per cent of 
water. The oil so expressed is of good color, but contains more 
mucilage and less of the glyceride element. 

The loss in the manufacture of the bisulphide is about 50 per cent 
of the charcoal and 17 to 18 per cent of the sulphur. 

The use of the bisulphide of carbon for a paint vehicle is more for 
the cheaper grades of roofing or color paints and for wooden struc- 
tures of minor importance than for the better grade of house or ferric 
paints, though in many of the latter it is used freely, judging from 
the odor. Its use is simply &s an adulterant of the oil and to cause a 
quick drying of the paint, and wherever used it may be considered to 
accompany a cheap oil, and the grade of the pigments mixed with it 
will in general be as low as the vehicle. 

Like benzine drieis, it sensibly lowers the tempei^ture of the 
surface of the body being covered, and in cool or damp locations this 
reduction is often enough to reach the dew-point and cause a sweat- 
deposit on the surface of the paint, causing it to peel. 

There are special grades of carbon-blacks, or bisulphide-of -carbon 
paints, under many trade-marks, specially noted in trtide literature 
for their excellence as coatings for brine, ammpnia, refrigerating and 
brewers' tanks and barrels. In some of these the paint appears to give 
good results. Nearly all of these paints are simply asphalt or natural 
bitumen, refined more or less, and a bisulphide-of-carbon vehicle con- 
taining little if any linseed-oil. They evaporate quickly and leave 
the bitumen coating behind, and probably coat the surface more thor- 
oughly than is possible to apply bitumen hot or in any other manner. 

For painting galvanized iron, the bisulphide of carbon appears to 
be of merit; at least the coatings containing some amount of bisulphide 
either as the principal vehicle or as a drier do not peel as readily as 
oil paints of similar color and pigments. This favorable point is 
more marked in the case of the brownish-black or full-black paints, 
and is probably due to the bisulphide element dissolving the greasy 
coating of the sal-ammoniac soap, that forms on the surface of the gal- 



BISULPHIDE OF CARBON. 205 

vanized sheet in the process of galvanizing. The presence of this 
soapy coating prevents the oil-paint coating from bonding to the 
metal, and it dries as a loose skin, peels easily, sometimes before 
it is dry. 

The bisulphide being a solvent of all semi-glutinous substances 
loosens up this soap and incorporates it into the mass of the coating, 
and the quick evaporation of the bisulphide leaves it there. 

There are many instances on record of the disastrous results upon 
the health of the painters who spread bisulphide-of-carbon mixtures. 
A noted one is its use with maltha (a mineral bitumen) for the internal 
and external coatings of a number of miles of steel-riveted water-pipe 
mains, where the application of this mixture was attended by the 
disability, insanity, and death of a number of the painters. Its use 
as an anti-corrosive coating for protecting miles of water-pipes was 
wholly experimental, without a single record on which to base such 
an application of an untried material, and especially one known to 
be decidedly inferior and uncertain for minor purposes. Had a gill 
of this maltha paint been spread in a room where the Board of Water 
Commissioners and their Engineering Staff held council over the pro- 
tection of water-pipes from underground corrosion, all the subsequent 
injury to the painters and expense of application and removal could 
have been avoided. Another coating was substituted for the maltha, 
but not before a number of miles of the water-mains had been laid 
and covered in with no better protection against corrosion than that 
which could have been had with a coating of boiled skimmed-milk 
glue. 

In the open air bisulphide mixtures can be spread without material 
danger or discomfort to the painters, but they have not a single ele- 
ment of protective value that warrants their application to any 
ferric structure of magnitude. They should only be spread on those 
of minor character, where the corrosion or decay is of no material 
importance, and the question of the cost of the coating and its tem- 
porary appearance governs. 

Frequent analyses of bisulphide paints show about 50 per cent 
of a low-grade resin, bitumen, and lampblack, for the pigment, with 
barjrtes or siUca added to give weight. Bisulphide-of-carbon coatings 
brush out easily and spread over a large area, as the vehicle is very 
thin compared with that of a linseed-oil paint. This feature of itself is 
against their protective quality. In such cases the pigments are but 
thinly covered or embedded in the vehicle, the quick dr3ring of which 



206 BISULPHIDE OF CARBON. 

by evaporation leaves a porous, crumbly mass with hardly any bond 
between the atoms of the pigment or to the covered surface. The 
dried coating soon shrinks in mass, cracks finely, is easily rubbed off 
by the hand, and requires to be wholly removed before repainting, 
even to apply another coating of the same compound. 

Analyses of dried bisulphide-of-carbon coatings show from 5 to 10 
per cent of sulphur, certainly not a suitable substance to recoat with 
any linseed-oil paint, unless the prompt peeling of it is desired. For 
ferric coatings, sulphur in any form or amount either in the pigment 
or vehicle is to be avoided. 

A new process for the manufacture of carbon bisulphide by an 
electric furnace has lately been developed and patented in the United 
States by Edward R. Taylor, of Penn Yan, N. Y.* Whether the new 
process will supersede the old or burning-charcoal process remains to 
be commercially demonstrated. Its many uses in the arts outside of 
paints will always cause it to be in demand. Its present price of 4 
to 4J cents per pound, equal to 42 to 45 cents per gallon, leaves nothing 
to recommend it as a substitute solvent or drier for turpentine in a 
paint. 

A French chemist, M. La Roy, has suggested an improvement 
in bisulphide of carbon as a substitute for turpentine in paints and 
varnishes. It is the chloride of carbon, or more particularly, the 
tetrachloride of carbon, CCI4. Its characteristics compared with 
turpentine are interesting. It is a colorless, limpid fluid, specific 
gravity, 1.56 or 13 pounds per gallon; boils at 170° F., being more 
volatile than turpentine, having an aromatic, pungent odor, is soluble 
in alcohol and ether, and dissoluble in water. The fluid is not in- 
flammable^ and dries quicker than turpentine. It can be mixed in 
all proportions with all of the usual paint solvents, including the 
bisulphide of carbon. Varnishes made from it are exceptionally hard 
and brilliant. 

In comparison, turpentine (CioHjg) has a specific gravity of 0.86 to 
0.88=7.176 to 7.343 pounds per gallon, boils at 160° F., and is quite 
inflammable both in fluid or vapor. The tetrachloride flattens the 
gloss in oil paints the same as turpentine, but adds weight to the paint. 

Tetrachloride of carbon f is produced : First, by the action of 



* " Carbon Bisulphide in the Electrical Furnace." Described in the Electrical 
World and Engineer, also in the American Gas Light Journal (N. Y.), Jan. 6, 
ld02, p. 11. Illustrated. 

t Watts's Dictionary' of Chemistry, Vol. I, p. 765. 



BISULPHIDE OF CARBON. 207 

chlorine on marsh-gas. Second, by the action of chlorine on chloro- 
form exposed to the sunlight. Third, the probable commercial proc- 
ess of manufacture, by the action of chlorine on the disulphide of 
carbon; the reaction being, CS,+ 4Cl3=CCl4+2SCl3. Chlorine, satu- 
rated with the vapor of the sulphide of carbon by passing it through 
the Hquid, is passed through a red-hot tube or retort filled with pieces 
of porcelain, the outlet of the retort being connected to a receiver 
packed in ice. The condensed yellow mixture of tetrachloride of 
carbon and chloride of sulphur thereby obtained is slowly added to an 
excess of potash lye or milk of lime, the mixture being agitated from 
time to time and afterward distilled. The tetrachloride of carbon 
passes over, mixed with some of the sulphide of carbon. If too much 
of the sulphide has been mixed with the chlorine, or if the decomposing 
heat has not been strong enough, the sulphide of carbon can be re- 
moved by leaving it for some time in contact with the potash lye. 

No estimated cost of the tetrachloride product is at present 
given, but its field of usefulness in the manufacture of paints and 
varnishes, also as a special drier, is favorably indicated from the few 
trials and experiments thus far had with it. 



CHAPTER XXI. 

JAPAN DRIERS. 

Japan driers or japans vary greatly in their composition and 
are very erratic in their action as drying agents. Specimens from 
the same manufacturer, taken from stock at different times, are 
widely different in drying qualities, while any attempt to classify the 
japans of different manufacturers is one of the vexations of the 
master painters. Probably a good rule for painters to follow in the case 
of japans is, when one has been found to suit them, to lay aside a 
sample of it to compare with all future supplies, and to stick to that 
manufacturer and brand just as long as it comes up to the mark. 

The general composition and process of manufacture of japans 
are: Gum shellac is cooked with linseed-oil in a varnish kettle until it 
becomes thick and partakes of the nature of a varnish. Litharge 
and other substances are added to quicken the drying of the resulting 
product. When the mass has cooked do^Ti to a thick substance 
called a "pill," it is allowed to cool and then thinned down with tur- 
pentine. Japan is a light-colored brownish-yellow liquid of about 
the consistency of varnish. A thin surface of it dries in from 15 to 
20 minutes. The care exercised in the manufacturing process and 
the purity of all the materials used, affect its quality, and are the 
cause of such erratic results from its use. 

The reputation of a japan or varnish manufacturer counts for 
much, but it does not always ensure a good article, if the price governs 
the selection. 

Formulae for japans are numerous and are trade secrets. The 
following are representative samples: 

One gallon cold-pressed old linseed-oil, J pound of D, C. or L. C. 

gum shellac, i pound gold litharge, i pound burnt umber, i pound of 

red lead, 6 ounces sugar of lead. Boil together with constant stirring 

for 4 hours, or until all of the ingredients are dissolved. Remove 

from the fire, and when cool add 1 gallon oil of turpentine; stir well 

while it is being added. 

208 



JAPAN DRIERS. 



209 



A cheap japan: Mix 4 gallons pure linseed-oil, 4 pounds each of 
litharge and red lead, 2 pounds of powdered raw umber. Boil slowly 
for 2 hours and add by degrees 7i pounds D. C. gum shellac, and 
boil i hour longer or imtil the ingredients are well mixed. Add by 
degrees 1 pound powdered sulphate of zinc, and when nearly cold, 
stir in 7 gallons of spirits of turpentine. 

The "Bung-hole Drier" formiike are as numerous as the oil 
compounders. The following represent a few of the compounds used: 

Lead Oils. 

Linseed- or nut-oil 1 gallon. Linseed- or nut-oil 1 gallon. 

Litharge 1 pound. Litharge } pound. 

Sugar of lead J pound. 



Manganese Oils. 

Linseed-oil 1 gallon. * Linseed-oil 

Potassium permanganate, 100 grains. Pure hydrated protoxide 

of manganese 



1 gallon. 
J ounce. 



Manganese and Lead Oils. 



Linseed-oil 1 gallon 

Umber 5 ounces. 

Gold litharge 5 ounces. 

Red lead 5 ounces. 

Linseed-oil 1 gallon. 

Permanganate of potash. . 4 ounces. 

Acetate of lead 4 ounces. 



Linseed-oil 1 gallon. 

Borate of manganese 1 ounce. 

Acetate of lead 1 ounce. 

Linseed-oil 1 gallon. 

Manganese protoxide hy- 
drate 1 ounce. 

Red lead or litharge 1 ounce. 



See also Boiling Oil, Chapter XXIII. For the effect of different 
driers upon linseed-oil, see Thorp's experiments, same chapter. 

The following is an extract from a Report of *' Test on Liquid 
Driers,*' read at the Sixth Annual Convention of the Master Painters 
and Decorators' Association of the United States, held in Detroit, 
Mich., on Feb. 11, 12, and 13, 1890. 



1 part Azote drier (trade-mark) to 1 part 
" " " " " " 5 parts 

a a it a a " 10 parts 

a a u it u ii 15 parts 



Raw 
Linseed-oil 

Dries in 
Hrs. Min. 



1 
2 
3 
4 



50 
35 
40 
30 



White 

Lead 

Dries in 

Hub. Min. 



1 
2 
3 
5 



60 
50 
50 
00 



Vandyke 

Brown 

Dries in 

Hrs. Min. 



2 
4 
6 




15 
20 
15 
40 



Lamp- 
black 
Dries in 
Hrs. Min. 



1 
2 
3 
6 



50 
20 
45 
20 



210 JAPAN DRIERS. 

Brown japan should mix well with raw linseed-oil in any pro- 
portion up to 15 per cent, and should stay mixed for at least 6 hours 
without showing sediment or separation^ called "curdling." 

When applied in a thin film to a clean^ dry piece of glass placed 
in a vertical position, the japan should be dry to the touch in about 
2 hours, and should dry hard without becoming brittle in 6 hours. 

The so-called concentrated driers are made by heating linseed-oil 
with lead and manganese salts or oxides in excess, until the product 
becomes viscous, like a sticking-plaster or birdlime. Liquid driers 
are concentrated driers, thinned out while hot with naphtha or spirits 
of turpentine. When applied in a thin film to glass and placed in a 
vertical position, they should be dry to the touch in 2 hours, and harden 
in about 8 hours. After 48 hours the drier should not rub off in the 
form of a fine powder when the finger is rubbed briskly over the sur- 
face. Liquid driers should mix freely with raw or boiled linseed-oil, 
turpentine, or benzine in any proportion without showing clots or 
precipitate after standing 48 hours in the open air. 

Inferior liquid driers can be recognized by the odor of benzine 
when the sample is slightly wanned ; by the powdering of the hardened 
film when rubbed by the finger, and by the rapid evaporation when 
exposed to the air, with consequent separation of the ingredients. 

The quality of a japan depends as much upon its cooking as upon 
the quality and kind of the materials in its composition. Too high a 
heat or too long exposure to the heat frequently spoils it. 

Gum is added by some manufacturers of japans to harden the oil. 
This, while causing the japan itself to dry more rapidly, reduces its 
power to dry an oil paint. Gum is a very uncertain substance in the 
formula of japan manufacturers. It may mean the spruce cud of 
the schoolgirl, common resin, or the best grade of the fossil resins, over 
thirty in number, with many varieties in each number. 



CHAPTER XXII. 

FLAX-PLANT AND LINSEED. 

Linseed is the seed product of the Linum usUaiissimum, This 
plant is a native of India or Eastern Asia, and its cultivation has 
existed from the earliest ages, distinct evidences of its existence 
during the Stone Age being preserved to the present day in the rough 
and worked flax made into bundles, found in the lake dwellings of 
Switzerland. 

It is mentioned in the book of Exodus as one of the products of 
Egypt in the time of the Pharaohs. Among the plagues of Egypt, that 
of hail destroyed the flax and barley crops, "for the barley was in the 
ear and the flax was boiled" (Exodus ix. 31). 

Pharaoh "arrayed Joseph in vestures of fine linen" (Genesis xli. 
42). 

Solomon purchased linen yam in Egypt and Herodotus speaks of 
the great flax trade of Egypt. 

Numerous pictorial representations of the cultivfiCtion and prep- 
aration of flax are sculptured on the walls and tombs of Thebes, 
showing the varieties of flax in the red and white flower, the manner 
of pulling, retting, and hachelling as practised when Jacob dwelt in the 
land of Goshen; and, except in some minor particulars, or in certain 
favored locations, are precisely the same as practised at the present 
day. 

The crushing of the seed in a mortar, grinding it on a stone slab 
by a muller, the pressing out of the oil with stones, the seed-bag, 
the burning lamp showing that the ancients knew the value of heat 
to aid in the extraction of the oil, and the painter with his bristle brush 
and paint-pot is also delineated. 

Flax is more extensively and successfully cultivated in Belgium 
than in any other part of Europe, that raised in East and West Flan- 
ders (the Coutrai flax) being the most valuable of the world's crop 
of this fibre. It is used in the manufacture of Brussels lace. The 

crop often exceeds in value the land on which it is raised, bringing 

211 



212 FLAX-PLANT AND UNSEED. 

J500 to $750 per ton, while the ordinary fibre crop brings $200 to 
$400. 

Prof, Hodge's (of Belfast) experiments with 7770 pounds of dried 
flax yielded the following resulta: 

1946 pounds of bolls, which furnished 910 pounds of seed. The 
6824 pounds (52 per cent) of flax fibre, lost in steeping 1456 pounds, 



Fio. 32. — Jerusalem flax-plant blossom. It growB wild in Palestine, covering 
lajge areas around Jerusalem. (Blue flower,) 

leaving 4368 pounds of retted stalks, and from that 702 pounds of 
finished fibre were produced. The weight of fibre was equal to about 
9 per cent of the dried flax stalk with the seed-bolls, 12 per cent of 
the bolted straw, and over 16 per cent of the retted straw. 

By Schenck's (American) method, 100 tons of the dried flax straw 
gave 33 tons of bolls with 27.5 tons' loss in steeping; 32.13 tons were 



FLAX-PLANT AND LINSEED, 213 

separated in scrutchings, leaving 5.9 tons of finished fibre and 1.47 
ton of tow and pluckings. 

Generally two bushels of linseed are sown per acre, and the 3rield 
in finished fibre is from 600 to 800 pounds, the market price of which 
is about 12 cents per pound. The yield of seed is from 8 to 10 bushels 
of 52 pounds, and is graded and classified as to quality and condition 
as closely as any of the grains. The crop is very exhausting to the soil ; 
potash and phosphoric acid are the chief ingredients that the soil 
requires to produce a good crop of either the fibre or seed. It requires 
from 400 to 600 pounds of mineral or phosphate fertilizers per acre, 
beside barnyard and other manures, to keep the soil in condition, 
and then only two or three crops can be raised in succession, when 
other crops must be substituted for from 5 to 8 years. 

New England formerly raised large quantities of flax for the 
fibre, but the advent of cotton manufacture soon displaced flax cul- 
ture, and this, with the exhaustion of the soil and absence of phosphate 
fertilizers, caused an abandonment of the flax crop in that part of the 
United States, early in the past century. 

America furnishes about one-fourth of the world's supply of 
linseed-oil. The crop of linseed for the years 1900-1901 was from 
16,000,000 to 17,000,000 bushels. The average yield of oil was 18| 
pounds of oil per bushel, or 2.465 gallons of 7i-pound oil; equal to 
40,000,000 to 42,000,000 gallons. In general, the American crop is com- 
paratively free from the adulteration of the wild mustard and other 
acrid seeds that render the oil-cake almost valueless for a cattle food. 
Though, in this respect, it is better than most of the foreign seeds, 
it is, however, the practice for many seed-crushers to add the screen- 
ings from the linseed and grain elevators to their linseed in the crushers, 
and this not only furnishes a bitter oil-cake but a poorer oil. 

The American linseed crop is now chiefly produced by the North- 
western States, where the rich prairie soil is favorable for a heavy seed 
crop without much fertilization. The fibre in these States, from its 
distance to market and the difficulty of preparing it, is of minor import- 
ance, and the plant is generally allowed to fully ripen before harvest- 
ing, the flax being burned, like the straw from the wheat-fields, to 
get rid of it. Duluth and Chicago are the commercial centres for 
the distribution of the Western linseed crop, the yearly production of 
which is not clearly determined at this date (February, 1902) owing 
to the incomplete state of the last United States Census. 

The Argentine Republic is the greatest flax-growing country in 



FLAX-PLANT AND LINSEED. 



the world. Flax-growing was begun in Argentina nearly a hundred 
yeais ago, but not until about 20 years back was any attempt made 



Pig. 33. — Flai-jdant — flower, seed-vessel, and seed, 
Ite flower is blue. Fig. 1 represents s, flower leaf or petal; there are five to 
each flower, which is of a verv re^lar and perfect kind, having five petals, five 

Sistils, five stamene, five sepals. Figs. 2 and 3 are sepals, or cup leaves, to the 
ower; Figa. 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 ten seeds. 
The stamens fertilize the pistils, the pollen falling upon the top of the pistil, or 
probably carried there by some busy liee. Within each of the pistils (not to 
speak exBctlv) 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 Flga. 
g, 10, 11, and 12. 

to raise it to the proportions of a national industry. In 1881 some 
67,000 acres were planted in flax in the province of Buenos Ayres. 
The success of the venture led to wider planting in that province. 



UNSEED: SOURCES OF SUPPLY. 215 

and in Coidovs, Entre Rios, and Santa F6. To-day the crop is one 
of the moBt important in the country, and surpasses in magnitude 
that of any other land. 

The plant is grown only for the seed, and as soon as the latter is 
secured the straw is burned. An average of 1000 poimds of seed is 
raised on an acre, and in some cases the yield is 2000 poimds. The 
e3q)ort of flaxseed from the four provinces named amounts to 500,000 
tons a year, which is one-half the entire product of the world and 
equals 54,200,000 gallons of oil. Not more than 20,000 tons are 
retained for domestic use, and there appear to be no linseed-oil mills in 
the country, as all the oil used there is imported. One wonders what 
the effect upon the markets of the world might be if Argentina should 
export linseed-oil and cake instead of raw flaxseed, and could transform 
the straw into linen thread and cloth instead of burning it. 

Ireland, England, Belgium, and Central Europe raise the best 
flax for fabric purposes, but seeds gathered from these sources being 
unripe, furnish poor, watery oil. 

Russia has a large acreage of flax for seed purposes and furnishes 
about one-sixth of the world's supply of linseed, the yield being 
about 8 bushels of 56 pounds to the acre; the flax fibre is of minor 
importance, being woody and subject to great waste in preparing it for 
fabric. 

Russian seed is exported for seed purposes as well as for oil extrac- 
tion. In Russia hempseed is sown with the flaxseed, and comprises 
nearly one-tenth of the seed crop, but as this seed furnishes a siccative 
oil, it is not an objectionable adulterant, such as the seeds from the 
rape, colza, mustard, and many other non-drying oil-seeds, called 
"flax-dodders." The adulteration from these acrid seeds is so great 
that the waste product in the form of oil-cake, formerly a valuable 
cattle food, is now so strongly impregnated with the biting taste of 
these seeds, that cattle refuse to eat it, and it is now used for fuel or 
fertilizing purposes. 

India furnishes about one-eighth of the world's supply of linseed. 
It is grown as a mixed crop for the seed only. The India flax-plant 
has been deteriorating for over 200 years, until it is now an inferior 
shrub from 12 to 16 inches high. The climate is favorable for the oil- 
producing quality of the seed. The white-flower plant produces about 
2 per cent more oil than the blue-flower variety, also a sweeter and 
softer oil-cake. The edges of fields devoted to other crops are sown 
with linseed for seed purposes, which is allowed to fully mature before 



216 QUALITIES OF LINSEED. 

gathering, the ordmary linseed crop being harvested just before the 
seed has fully matured, and while it contains more water than if fully 
ripened. 

Rape-seed is sown in large quantities with the Unseed. Its yield 
of seed and oil is very large, and when refined it passes as colza-oil 
from coleseed. These seed-oils are used for burning, lubrication, 
and in the manufacture of india-rubber articles, because of their 
non-drying qualities. India is very prolific in oil-bearing seeds; the 
mustard and many other acrid seeds grow wild, are very rich in oil, 
and all are freely used to adulterate linseed to an admitted amount 
of 10 per cent and possibly 15 per cent more. 

The quahty of the flax, also of the seed, varies quite as much as 
any crop of grain or vegetables, according to the locality in which they 
are raised, the soil, weather, and other influences affecting the fibre 
or oil, and the crop is quite as exhaustive to the soil as wheat or com. 

Samples of linseed grown in various parts of the world and aver- 
aged from a collection of ripe seeds weighed from 48 to 52 pounds 
per bushel, and the yield of oil was quite as variable, viz. : 

Gallons op 7J-pound Oil per 112 Pounds of Seed. 

Best Odessa seed 15 to 16 gallons. 

Archangel " 18 "19 

Good commercial seed 15.5" 16 " 

East Indian seed 17 " 16.5 " 

Sicilian " 16 " 16.5 " 

General results by a large crusher for all seeds 14 " 17 " 

American seed, 52^ pounds per bushel, gave 26.55 per cent of oil, 
or 13.87 pounds. 

Linseed in its dry^ state, as analyzed by Dr. Ure, contains: 

Oil 11 .265 per cent. 

Wax 0. 146 " 

Soft resin 2.488 " 

Resinous coloring matter . 550 " 

Yellowish coloring matter analagous to tannin .926 " 

Gum 6 . 154 " 

Vegetable mucilage 15 . 12 " 

Starch 1 .48 " 

Gluten 2.932 " 

Albumin 2.782 " 

Saccharine extractive 10 .884 " 

Enveloping material, including some vegetable mucilage 44.382 " 

99.109 " " 
Other substances and loss 891 includ- 
ing free acetic acid, some acetate, sulphate, and muriate of potash, phosphate 
and sulphate of lime, phosphate of magnesia and silica. 






LINSEED: ANALYSES OF. 



217 



Analyses of linseed by Meurein {Journal of Pharmacy [3], XX, 96) : 

Gum and soluble salts 14 per cent 

Softi resin and fixed oil 1 

Matter insoluble in water but soluble in ether. ... 4 

Water 2 

Soft resin and fixed oil 6 

Matter insoluble in water but soluble in ether. ... 12 

Matter soluble in water 3 

Water 2 

Fixed oil 30 

Matter insoluble in water but soluble in ether. ... 18 

Matter soluble in water 3 

Water 5 



Analyses by Anderson:* 

Albuminous substances 24 . 44 per cent 

Gum and cellulose 30.73 " " 

Oil 34.00 " " 

Ash 3.33 " " 

Water 7.50 " 



tt 
it 


tt 


Episperm. 
21 percent. 


11 


" J 




ti 


M 




tt 
tt 


tt 
tt 


Endosperm. 
' 23 per cent. 


tt 


tt 




tt 


It 




tt 


tt 




tt 


tt 


' 66 per cent. 


tt 


tt 






100 per cent. 



tt 



â–  100 per cent. 



Way's analyses of 33 samples of linseed from various countries: 

Nitrogen 3.3 to 5.28 percent. 

Fat 34.70" 38.42 " " 

Ash 2.68" 5.64 " " 

Water 8.51" 12.33 " " 

Way's analyses, ditto of the oil-cake from above samples: 

Nitrogen 3 . 92 to 5 . 25 per cent. 

Fat 6.60 " 15.32 " " 

Ash 5.45 " 22.66 " " 

Water 6.56 " 10.26 " " 

Albximinous substances. 25.00 " 36.00 " " 

The general composition of all siccative oils is: 

Carbon 77.40 to 76.00 per cent-j 

Hydrogen 11.30 " 11.10 " " ^Konig. 

Oxygen 12.70 " 11.50 " " 3 

Linseed f also contains a large quantity of mucilage deposited in 
the outer layers of cells of the epidermis, which swells up on macerating 



♦ " Analyses of Linseed." Highland Agricultural Society Journal (New Series), 
Number 69, p. 376. 

t Schmidt. Amer. Chem. Phar., II, 26. 



218 



LINSEED-OIL: ANALYSES OF. 



the seed with water, sufficient to burst the cells. One part of linseed 
boiled in 16 parts of water yields mucilage enough to be drawn out 
into threads, and forms a dark-colored spongy mass when dry. This 
crude mucilage contains in addition to the true vegetable mucilage, 
legumin, albumin, and an organic acid, probably malic acid; also 
ash constituents, chiefly lime, potash, and iron, partly as phosphates, 
and partly united in the ash by carbonic acid. Linseed mucilage 
precipitated by alcohol gives 11 per cent of ash containing 4 per cent of 
carbonic acid. 

Linseed-oil has a specific gravity of 0.928 to 0.953, or 7.743 to 
7.952 pounds per United States gallon. The oil from an average 
quaUty of ripe seed extracted by various processes contains: 



Cold Process. 

Carbon 75 . 17 per cent. 

Hydrogen 10.98 

Oxygen ia.85 



Hot 



(( 



tt 



« 



tt 



Carbon 78 . 11 per cent. 

Hydrogen 10.96 " 

Oxygen 10.93 



tt 



tt 



it 



100.00 



tt 



tt 



100.00 



« 



tt 



The carbon-disulphide process gives more oxygen and less carbon. 
The oil extracted by the naphtha and percolating process does not 
show any material difference in the quantity of the oil, but is thought 
to give a quicker drying oil than by the old or cold-drawn process. 
But, however, it leaves some of the glucerides of the oil in the oil-cake 
as well as some of the albumin. 

The glucerine in the oil in the form of gluceride or other ethers is 
needed in the change of the fatty acids to form the soap compounds 
that give the binding quaUty to the oil. The albuminous substances 
are organic and are subject to decomposition, and constitute "the 
drops" or "mucosities" that the boiling process removes, or if the 
oil is used in its raw state, the driers added are intended to affect them 
so that they may be oxidized and dried. 

Unripe linseed or the seed from flax raised for the fibre (the condi- 
tion that furnishes a large part of the commercial linseed-oil) contains 
5 to 8 per cent of water. 

The yield of oil from the different classes (red, blue, or white flower) 
of linseed varies from 20 to 33 per cent of the weight of the dry seed. 
The classification of American Knseed, in regard to its quality, is 
established by the Board of Trade of the city of Chicago, as follows: 

No. 1. The minimum weight per bushel shall be 51 pounds, the 



LINSEEDjOIL: QUALITIES OF. 



219 



maximum quantity of field-stock, storage, or other damaged seed 
not to exceed 12^ per cent. 

No. 2. The weight per bushel shall be 50 pounds, the damaged 
seed not to exceed 25 per cent. 

No. 3. The weight to be not less than 46i pounds per bushel; 
the damaged seed, not to be in excess of 20 per cent, is graded *' Re- 
jected.'* 

No. 4. No grade. Seed comprises all damp, mouldy, warm seed 
or those in a heated condition and unfit for temporary storage. All 
seed that is burnt, smoky, or intermixed with burnt seed is posted 
as "Burnt or Smoky Flaxseed.". 

All sales of flaxseed are made upon the basis of pure 9eed; that is, 
seed tendered for contract deliveries may carry impure, damaged, or 
foreign seed matter, but must contain the sale-quantity of pure seed as 
given, and for such pure seed only shall pa3anent be required. 

Linseed shields by the 



Cold process of extraction, about 20 per cent of oil. 
Hot process " " " 27 

Carbon-disulphide process, " 33 









tt tt 
tt tt 



Or from 15 to 18 pounds of oil per bushel of No. 1 commercial linseed 
by the most improved processes of extraction. 

The average results in oil of a number of samples of the following 
substances, extracted by filtration in 100 parts, are as follows: 



Linseed 27.15 percent. 

Hemp-seed 25.87 " " 

Poppy-seed 49 . 40 



tt 



tt 



Wahiuts 50.06 percent. 

Almonds 62.41 " 

Grape seeds 17.95 " 



tt 



it 



Linseed-oil corresponds to the formula (Mulder) Cj^jsOj or 
CoH .4O3. It is an hydride of Unoleic acid (CnHjeOj). Specific gravity, 
0.9266. Linoleic acid is a faint-yellow, limpid oil, insoluble in water, 
and does not solidify at 18° F., and has both the nature of an oil and 
a resin. It decreases continually in weight for 90 days, losing from 
6 to 8 per cent. 

Linoleic acid is peculiar to all siccative or natural-drying oils, 
and when fully oxidized by exposure to the air, forms oxylinoleic 
acid (CjjHmOs) or linox3m (Mulder). 

Hazura and Bower (Monatsch.y Vol. I, p. 469) found that the rate 
of oxidation and consequent hardening of linseed and other siccative 
oils depended upon the ratio of linoleic and linolenic acids present. 



220 LINSEED^IL: QUALITIES OF. 

The linoxyn formed is insoluble in water, dilute acids, alcohol, or 
ether, and is heavier than water. 

Hazura describes linseed-oil as formed of linolic, linoleic, and 
iso-linolenic acids, and that a high proportion of the two latter acids 
is characteristic of this oil. 

Linseed-oil is composed of drying oil, 80 parts, and non-drying or 
fatty oils, 20 parts. Of the latter, 8 parts are glycerine ether, the 
volatile element of the oil, that in the chemical changes among the 
oil acids by the absorption of oxygen in the process of drying, is 
absorbed or lost in the change to linoleic acid, with a direct loss in 
weight of the new oil compound. 

The fatty acids in linseed-oil are: 

Margaric acid C^Hy^O,; specific gravity .810 

Palmitic or benic acid CjeH^O, " " 0.809 

Oleic acid C.^Hg^Oa " " 0.808 

Stearic acid Cj^HaoO, " " 0.806 

Margaric acid is considered simply as a mixture of stearic and 
palmitic or benic acid of identical composition (CigHj^Oj). 

The composition and specific gravities of all the fatty oils vary 
but little, and the influences that affect one affect all. 

Oleic ether (CjoHjjOa), specific gravity^ 0.807, associated with the 
fatty acids, dissolves all solid fats, stearic, palmitic acids, etc. 

The glycerides or glycerine ethers have the characteristics of 
ethylin (CgHuOa) ; they are the most volatile of the group, and form a 
component part of the fatty elements in all siccative oils. The heavy 
odor recognized when a burning tallow candle is blown out is due 
to the glycerine ether that comes from the smoldering wick. A 
heat of 170® F. is adequate to dispel or to cause an absorption of the 
glycerides into the fatty acids in the oil to form the insoluble soap com- 
pound. The albuminous substances in the oil coagulate at about 
160® F., and in the steamed or hot process of extracting linseed-oil, 
the meal is cooked at 190® to 200® F. to prevent the albumin from 
flowing out when pressed. 

These changes indicate the merits of a low and long-continued 
heat in the process of boiling oil, instead of the quicker and more 
energetic changes due to higher temperature of the oil, and will be 
referred to hereafter. 

The siccative oils of commercial importance are 20 in number, 
the principal of which are linseed, poppy, hemp, walnut, sunflower, 



SICCATIVE AND NON-DRYINO OILS. 



221 



grape, Scotch and silver fir, and spruce. The specific gravities of 
the whole number (20) range from 0.9202 to 0.9358, varying so little 
that the hydrometer-test is of little moment to detennine their char- 
acter. The vegetable non-accative oi)a of commercial importance 




Fio. 34.— Oil-seeds. (Enlarged.) 



are 76 in number, the principal ones being the castor (specific gravity, 
0.964), olive, cottonseed (specific gravity, 0.9306, almost identical 
with linseed), remn, almond, beechnut, horse-chestnut, hazelnut, 
peanut, croton, sesame, colza, rape, mustard, with specific gravities 
ranging from 0.913 to 0.942. They mix thoroughly with each other 
and with the siccative oils, and the specific gravity of a pure llnseed- 
otl can be eauly counterfeited, even if its quality cannot be. Veg^ 



222 COMPOSITION OF SICCATIVE AND NONJURY INQ OILS. 

table oils additional to the above, used medicinally and for soap, 
burning, and food, are 85 in number. 

Vegetable oils, volatile and essential, number 134, or a total of 
315 non-drying oils available for the purpose of adulteration. Add 
80 animal and fish oils to the above, and it may be conjectured, the 
source from which a yearly supply from the whole world of 250,000,000 
gallons of linseed-oil and other siccative oils can supply a demand 
for about 350,000,000 gallons of paints, varnishes, japans, and other 
uses, including the wastes of manufacture. 

As before stated, if the flax is raised for the fibre, the seeds do not 
mature at the same time, hence they furnish a thin watery oil with a 
greater quantity of the albmninous substances, gum, sugar, and cellu- 
lose, called "mucosities," than the 54.44 per cent to 58.50 per cent 
found in ripe seed. Unripe or mildewed linseed is no more capable 
of furnishing a good oil than a green apple will make either good 
cider or good vinegar; an imripe grape, good wine; or grain or com 
harvested in the milk will make good bread ; and no amoxmt of juggling 
in the subsequent manipulations of the paint manufacturer can 
replace the simple operations of nature, or ripen her unripe products 
or produce them from S3aithetical compounds. 

The following is a comparison of the composition of a few of the 
principal oil-seeds: 

Siccative Oius. 



OU. 

Linseed 

Poppy-seed (black). 

^' " (white) 

Hemp-seed 

Walnut 



Specific 
Gravity. 



0.9352 
0.9270 
0.9285 
0.9307 

0.9288 



Carbon. 



77.40 
76.57 
77.20 
76.00 
76.65 
77.15 



Hydrogen. 



11.10 
11.41 
11.31 
11.30 
11.46 
11.73 



Oxygen. 



11.50 
12.02 
11.49 
12.70 
11.89 
11.12 



Vegetable Non-dryinq Oils. 



Cottonseed 

Rape (winter) 

" (summer).... 
Earthnut or Peanut 

Beechnut 

Castor 



0.9306 

0.9155) 

0.9165 y 

0.918 

0.923 

0.9639 




12.27 

9.92 

12.70 

11.89 

9.43 



COMPOSITION OF NONSICCATIVE AND NON-DRYING OILS, 223 



The grains yield oil similar in character to the other non-drying 
oils, viz.: 






e. .. 
leat. 



Barley and oats. . 
Maize and lupine. 

Peas and beans. . 
Potatoes and rice. 
Beets, etc 



Specific 
Gravity. 



i 
I 



Carbon. 



76.71 
77.19 
77.50 

to 
75.67 
77.40 

to 
76.00 



Hydrogen. 



11.79 
11.97 
11.96 

to 
11.43 
11.30 

to 
11.15 



Oxygen. 



11.50 
10.84 
12.78 

to 
10.69 
12.70 

to 
11.50 



The increase in oxygen in the grain oils is at the expense of the carbon. 
Miscellaneous Non-drtino Oils and Substances. 



Sperm-oil. . 

Resin-oil. . . 

Beef tallow 
Beeswax. .. 
Spermaceti. 



0.945 


78.90 


10.97 


(0.984 ) 






1 to [ 


79.27 


10.15 


(0.9910) 






0.938 


79.00 


11.70 




81.60 


13.90 




81.60 


12.80 



10.13 

10.58 

9.30 
4.50 
5.60 



The vegetable aromatic volatile oils from lemons, oranges, cloves, 
cinnamon, etc., are of the same chemical composition: Carbon, 88.25; 
hydrogen, 11.75. When they oxidize, it is at the expense of the 
carbon element. The diversities in their odor are due to a different 
arrangement of their atoms. 

Poppy-seed furnishes a cold-pressed siccative oil of excellent 
quality. Being an imported article its price prevents its use except 
in the finer color paints. The exceptionally good results attendant on 
the use of French zinc oxide are principally due to the use of this oil. 
It dries slower than linseed-oil and faster than walnut-oil. It does 
not remain sticky so long as linseed-oil, not being so fatty, and takes 
up less oxygen. It is used for the finer classes of varnishes. It con- 
tains 20 per cent of non-drying and 80 per cent of drying oil. Linseed- 
oil clarified by sulphuric acid has generally taken the place of poppy- 
oil. Walnut-oil contains 30 per cent of non-dr3dng and 70 per cent 
of dr3dng oil. The cold-process oil is light colored with a pleasant 
smell and taste; after a short exposure to the sunlight it becomes 
as clear as water. The hot-process oil is characterized by a deep color 
and unpleasant odor; it has more mucilage in it, and does not dry 
as well as the cold pressed. 



224 SPECIFIC GRAVITY OF OILS. 

Specific Gravities of Oils axd Fats 

Poppy-seed oil .9245 

Raw linseed cold-drawn oil 0.9299 to 0.932 

BoUed linseed-oU 0.9400 " 0.942 

Crude cottonseed-oil 0.9224 

Refined yellow ditto 0.9230 

Water-white ditto 0.9288 

Menhaden-oil (dark) 0.9292 

" " (light) 0.9326 

Tanner's cod-oil 0.9205 

Porgy-oil 0.9332 

Resin, commercial yellow 1 . 0700 

Resin-oil, first run .9835 

" " third run 0.9887 

" " other runs 0.9910 to 0.960 

CSrude Lima petroleum . 839 — 7 pounds per gaL 

Lard-oil 0.915 

Whale- or train-oil 0.925 

Sperm-oil 0.943 

Porpoise-oil 0.937 

Beef tallow 0.927 

Mutton tallow 0.938 



CHAPTER XXIII. 

BOIUNG LlNSEED-OlL, 

J. Nelson NeU's Experiments. 

The study of the composition and properties of the siccative oils 
has been but little advanced since Mr. J. Nelson Neil was awarded the 
Isis Gold Medal of the Society of Arts in 1832 for a paper read before 
that society on "Oil Boiling and Varnish Making" (published in the 
Transactions of the Society, Vol. XLIX, Part II), which treats so 
exhaustively upon the modes of manipulation, recipes, and precau- 
tions to be used, that the paper is the foundation for all subsequent 
accounts and modes of manufacturing varnishes. The additions 
and modifications which have been worked out since that time have 
not materially altered the processes of Mr. Neil, either in the better- 
ment of the quality of the varnish or the oil product, but relate more 
particularly to a shorter time and possibly less cost of manufacture 
than given by him, which comprised the application of direct fire to 
the kettles instead of the steam-jacket kettles and devices used by 
later experimenters and manufacturers. 

The substances that act catalytically part with some of their 
oxygen in the oil, and become to a certain extent deoxidized, and 
again coming into contact with the air either mechanically bllJwn 
through the combined mass of oil and driers, or by surface exposure 
in the settling-tank or barrel, recover their primal condition, and are 
ready to do the same work over again. It is the necessity for the 
reoxidation of the driers that causes the general adoption of the air 
agitation and steam processes now in general use for the boiling of 
oil and manufacture of varnishes, etc. These driers, without becom- 
ing materially altered, induce an alteration in the linseed-oil sub- 
jected to their operation. We may imagine this action as similar 
to that by which spongy platinum explodes a mixture of oxygen and 
hydrogen, or a platinum wire is kept red hot by the vapor of ether. 

226 



226 BOILING LINSEEM)IL, 

In properly prepared varnishes and paints from pure linseed-oil, 
the gain in weight after thoroughly diy is from 8 to 10 per cent of 
the vehicle, and it is heavier than water. Any paint that does not show 
a marked increase in weight in dr3ring can be set down as one of the 
hundreds of bastard compounds that masquerade under the guise of 
paint, with more resin, fish-oil, and hydrocarbon substances in its 
vehicle than linseed or other siccative oils. 

M. E. ChevreuPa Experiments. 

These experiments of Mr; Neil were followed by the experiments 
of Mr. M. E. Chevreul, who contributed a paper in 1856 to the Annales 
de Chime, corroborating Mr. Neil's deductions upon the drying of 
siccative oils, and by him clearly laid down, viz.: 

First. That it is the absorption of oxygen by the siccative oils 
and the change of the oleic, margaric, and stearic acids of which they 
are composed, and the chemical combination with each other in the 
presence of oxygen into the linoleic and linolenic acids, that is the 
cause of their solidification, which term he thinks more clearly defines 
the action of the oil than dr3ring, which in general may mean evapora- 
tion, which is a term adaptable to all liquid bodies, or rather indi- 
cates the removal of liquid from all bodies. This definition appears 
to be apropos to many of the latter-day cheap mixtures called paints; 
the difficulty experienced with some of which is not to have them 
dry in a reasonable time, but to have them keep liquid long enough 
to spread them at all. 

Second. That the oxidation of the oil is a chemical process and 
naturally inherent in itself. The action of heat, as in boiling, hastens 
th% drying or resinification of the oil by removing the water and 
mucosities. That all substances which can be used as driers must 
be such as are capable of parting with oxygen or dissolving in it; and 
being of themselves oxidizable in combination, they in that way 
increase its absorptive power. There is a class of driers (white cop- 
peras, for instance) which act catalytically, while mechanically sus- 
pended or in contact with the oil, and increase its oxygen absorptive 
power by their presence, but leave no increase of dr3dng power when 
withdrawn. 

Third. That manganese and litharge were the most powerful 
driers, and when used in excess and settled from the oil, they acted 
with greater power when used a second time. 



BOILING LINSEEDjOIL. 227 

Linseed-oil boiled 5 hours wUhovi drier ^ required 38 days to become 
thoroughly hard. When boiled the same period with new peroxide 
of manganese, it dried in 2 days. When boiled 6 hours with peroxide 
of manganese that had been used many times before, it required 
one-half day to dry firm and hard. 

Fourth. That manganese driers for exposed paints appear to be 
less durable than red lead or litharge driers. They harden quicker 
and become more brittle, from the harder character of the soap they 
contain, which is further developed in hardness by the heat of the sun. 

Fifth. The manganese-drier paints appear to peel more readily 
than red-lead driers, especially upon ironwork that .has but few 
points or irregularities of the surface to which the paint can adhere. 
Hence such a paint for an iron surface must remam in a measure 
softer and more elastic, or else it will be thrown ofif or peel by the 
changes in temperature and the expansion and contraction of the 
metal. 

Peroxide of manganese is electro-negative to iron and steel, and 
is noted for the freedom with which it imparts its oxygen to these 
metals. It is used in the manufacture of steel as a pui^ to bum 
out the impurities in it. 

Analysis op Dioxide of Manganese or Pyrolusitb (MnO). 

Red oxide of manganese 84 .05 to 87 . 00 per cent. 

Oxygen 14.58 " 11.45 " " 

Sesquioxide of iron 1.30 " 0.40 " " 

Alumina 0.30 " 0.00 " " 

Baryta 0.67" 1.20" " 

Calcium traces " 0.00 " " 

SUica 0.80 " 0.51 " " 

Water 5.80" 1.92" " 

Sixth. That linseed-oil heated until it lost one-sixth of its weight 
became thicker, unctuous, and viscid, and dried more readily, forming 
a tough, crude, turpentine-Kke mass, scarcely soluble in any other oil 
(printer's varnish). When linseed-oil is heated to 325° to 376® F. 
it will take fire and continue to bum without any further application of 
heat from without, until only tar or charcoal remains. 

Nut- and poppy-seed oils also possess this feature, which is some- 
times employed as a test of the purity of these oils. If these oils 
are adulterated to any great extent with fish, resin, or mineral oik, 
they will not continue to bum after ignition without a further addl* 
tkm of heat. 



228 BOILING LINSEED-OIL. 

Prof. Chevreurs experiments to show the effects of atmospheric 
gases on the drying of linseed-oil were as follows: 

Four panels of wood were painted on one side with white lead 
and on the other side with zinc oxide, the vehicle being raw linseed-oil. 

No. 1 was placed in a closed glass vessel exposed to carbonic-acid gas. 
No. 2 was placed in a similar vessel exposed to confined air. 
No. 3 " " " " . " " "free air. 

No. 4 " " " " " " "oxygen gas. 

The results were as follows: 

After 24 Houra. After 72 Houn. 

•^r •. r^ 1. . ' ( The white lead nearly set. Set but without adhesion to 

No. 1. Carbonic- \ ^u j 

, < the wood. 

^ ' ( The zinc oxide still fresh. Absolutely fresh. 

-kT « T • .^ J . \ The white lead nearly dry. Perfectly dry. 

No. 2. Lunited air. •{ ^, . ., * u * i. j u u 

{ The zinc oxide set but not dry. " " 



it u 



-^ . j The white lead nearly dry. 

'1 The zinc oxide set but not dry. " " 

No. 4. Oxygen gas. Both perfectly dry. 



i€ «r 



Professor Vincent's Experiments. 

Prof. Chas. W. Vincent's experiments in 1859 were upon the line 
of boiUng oil without the presence of driers, and that a high tem- 
perature was not necessary. The temperature used by Mr. Vincent 
was that due to steam at 40 pounds per square inch, 267° F., used in a 
steam-jacketed kettle in connection with mechanical agitation by 
revolving blades and a current of compressed air moderately heated 
by the act of compression. This process is that in general use at 
present in the manufacture of linoleum, and it is believed that the 
Germans practised this method many years preceding 1859. Pro- 
fessor Vincent's conclusions, drawn from experiment, were that air 
blown through the mass of heated oil is not as important a part of the 
process as many have assigned to it, and in reality effects nothing 
toward making the oil a drying one. He boiled linseed-oil with air 
alone, biU without drierSy for three days consecutively, keeping up a 
high temperature the whole time, and the resultant boiled oil re- 
quired precisely the same time to dry as the raw oil from which it 
was prepared. The body, however, had become so much increased 
that its consistency was more that of a varnish than an oil. In the 
oil subjected to the heat alone for the same time, without any air, 
except such as came to it in contact with its surface in the kettle, 



BOILING LINSEED-OIL. 229 

there was no such increase in its consistency as in the former case; 
the oil simply became more greasy, had less difficulty in penetrating 
the capillary tubes of paper, plaster, etc., than it previously had, and 
had decidedly less drying power. The oil that had been boiled with 
the air-blast was less greasy and had a greater consistency. Briefly, 
the surface exposure to the air and the heat secured a sufficient amount 
of body. The driers produced any required shade of color in the oil 
and reduced the time of drying from 3 and 4 days, for the raw oil, 
to 6 hours in the summer and 8 hours in the winter for the boiled. 
Boiled oil that is subject to long voyages at sea is apt to become fatty 
and not free working. This is due to the agitation it gets from the 
motion of the ship while it is under the increase of temperature due 
to the hold or cargo space in the ship, being, in fact, a long-continued 
low-temperature and agitation process of boiUng. The manufacturer 
^ards against this result by adding to boiled oil for shipment by 
sea, about one-fourth its volume of raw oil, the oil becoming brighter 
in consequence of the addition. Professor Vincent's steam-kettle 
process of boiling gives an oil of a lighter shade than the direct fire 
or high-temperature process. In both processes, acrylic acid (CgH^Oj) , 
a monobasic acid, is produced by the oxidation of acrolein. AcryUc 
or acroleic acid when purified, is a colorless liquid of a slightly empy- 
reumatic odor, lighter than water and mixable with it in all propor- 
tions. 

Acrolein (CjH^O) is the acid principle produced by the destruc- 
tive distillation of fatty bodies, resulting from the decomposition 
of glycerine (CgHgOa). Acrolein is a colorless, limpid, strongly 
refracting liquid, lighter than water, boils at 126° F. ; vapor density, 
1.897. Its vapor is so intensely irritating that a few drops diffused 
in a room are sufficient to render the atmosphere imsupportable. 
It bums with a clear bright flame, and dissolves in 40 parts of water 
and readily in ether. The solutions at first are neutral, but gradu- 
ally oxidize and turn acid in contact with the air. Under water it 
changes into a resinous substance (disacryl-resin) and the water 
becomes chaiged with acrylic, formic, and acetic acids. The vapor of 
acrolein passed through a red-hot tube is decomposed with the formation 
of water and charcoal. Its vapor is highly corrosive to iron bodies. 

Acrolein is developed by the decomposition of the mucosities 
in the raw linseed-oil under the influence of the boiling heat, and 
indicates the slower decomposition of these substances in a raw-oil 



230 BOILING LINSEED-OIL. 

vehicle for a paint. The slower process of solidification of the raw 
oil enclosing them, as it were, in a film of the vehicle, to develop later 
by decomposition into a destructive agent of the paint. The pig- 
ments in the paint may delay this decomposing action for a time, 
but cannot wholly prevent it, and in many cases hasten it. The 
removal or destructive change of these mucosities is absolutely neces- 
sary in baking japans, varnishes, japan driers, and linoleum; and if 
found detrimental there, it must necessarily follow that they are 
equally so in a paint oil. The decomposing point in a pure linseed- 
oil made from thoroughly ripe linseed is nearly 100® F. higher than 
in an oil made from green or damaged linseed. 

Professor Sacc's ExperimerUs. 

Professor Sacc's experiments in brief were, 2500 grains of oil 
boiled for 10 minutes only, with 30 grains each of litharge and red 
lead, and weighed after 24 hours' exposure to the atmosphere, the 
oil had lost only 60 grains. This sample increased 20 per cent in 
weight after complete resinification. A second sample boiled until 
there was a loss of 5 per cent in weight of the oil, the product assumed 
a molasses consistency, and did not resinify after 15 days' exposure 
to the atmosphere. A third sample boiled to a loss of 12 per cent 
became a caoutchouc-like mass that the atmosphere had no effect 
upon whatever. It was insoluble in alcohol, ether, chloroform, and 
bisulphide of carbon; boiling naphtha only dissolved traces of it. 
The only action which dilute acids had upon it was to extract a small 
quantity of the oxide of lead due to the driers. Hydrochloric acid 
dissolved it slowly, while concentrated sulphuric and nitric acid 
dissolved it rapidly, as they do all vegetable and animal oils and tis- 
sues. This substance was in fact identical with the product obtained 
by the modem process of boiling oil for the manufacture of linoleum. 

Linseed-oil submitted to a dry distillation (without boiling) gave 
off a white vapor (acrolic) from which was condensed a colorless oil 
(acrolic acid), having the odor of fresh bread, then expanded and 
yielded a distillate, a brown empyreumatic product; finally, a mass 
resembling jelly and caoutchouc remained. 

The SvlphuriC'Ocid Process. 

This process of boiling oil has been adopted by a number of manu- 
facturers. It furnishes an oil of light color which dries well and rap- 



BOILING LINSEEIU>IL. 231 

idly, mixes with all pigments without leading to any discoloration. 
Whites retain their purity of tone unchanged, but with all these points 
in its favor, the process cannot be recommended for the preparation 
of a vehicle to be used for a ferric protective coating. Briefly, the 
oil is first treated with a dilute sulphuric-acid bath (containing about 
30 per cent of sulphuric acid) which is agitated with the oil by the air- 
blast to dehydrate it, but is said to be not strong enough to carbonize 
it. After standing to allow the oil and acid to separate, the oil is 
run ofif into the usual steam-jacketed kettle heated to about 267° F., 
air is blown through the mass, while a solution of manganese linoleate 
in some hydrocarbon spirit (probably benzine) is added gradually 
during the process of heating and blowing. Cautionary care is re- 
quired not to add too miich of this material. It is the writer's opinion 
that this caution should extend to the point of not adding any manga- 
nese asociated with any hydrocarbon vehicle to the oil, and that pro- 
hibition should extend to the use of any sulphuric, nitric, or other 
caustic acid of any strength to the oil in its preliminary stage. It 
is almost impossible to clear an oil of either high or low specific gravity, 
of any acid of whatever strength of solution to which it may be ex- 
posed. More or less of the clarifying acid will be held in the oil either 
free or in combination with the water in the oil, that even a long 
washing with water, aided by an air-blast agitation, imU not remove. 
Kerosene, naphtha, gasolene, etc., are purified by treating with 
sulphuric acid and then thoroughly washed with water and a long 
agitation by the air-blast, and are then often found to contain acid 
enough to perforate the tin cases in which they are shipped. The 
slight improvement in the color of the boiled oil by this process is a 
very poor recommendation for its use. The same results are obtaina- 
ble by the ordinary steam-kettle process, using well-known mineral 
substances; for instance, the zinc and manganese salts will remove 
or throw down the mucosities, clear the oil to any desired shade, and 
cause it to dry promptly, while the water in the oil will be evaporated, 
naturally by the heat, and the dangers of the sulphur element avoided. 

Thorp's Experiments with Driers, 

The action of various mineral and metallic driers upon linseed- 
oil in the process of boiling to determine their effect on the color of 
the oil and the time of drying were made by Mr. Frank H. Thorp, S.B.* 

♦ Journal of Chemical Industry, Volume IX, p. 628, 1890. Reprinted in 
the Scientific American Supplement, No. 757, Volume XXX, June 5, 1890. 



232 BOILING LINSEEDjOIL. 

The oil experimented upon was in every case from the same barrel, 
and was a very light-colored, cold-pressed, Calcutta raw linseed-oil, 
specific gravity, 0.93. The weight of oil under test was in each case 
the same, 50 c.c. weighing 45.7 grms. The several samples were 
treated in glass beakers arranged in a sand-bath under temperatures 
from 200"^ to 300"^ F. In general, the temperatures from 230*^ 
to 275® F. gave the best results. The time of actual boiling was 
from li to 2i hours, and the percentages of driers varied from less 
than 1 per cent to 2 per cent by weight of the oil treated. Litharge 
furnished an almost colorless oil of firm film, drying in from 6 to 10 
hours. Lead carbonate, lead acetate, and lead borate, each furnished 
slightly colored oils with good films, drying in the order named, in 
10, 12, and 20 hours. Red lead, lead chloride, and lead tartrate 
furnished dark-colored oils of good films, drying in from 20 to 24 
hours. Red lead and litharge, 2 per cent of each, also the other lead 
salts mentioned, with larger percentages than two of each, gave dark- 
colored oik, all with firm films, drying in about 24 hours. 

Of the lead driers, litharge gave the best results both in color, 
film, and drying qualities. Care was necessary in the use of this 
drier, to not overheat the oil, thus deepening its color. The zinc 
salts, the acetate, borate, citrate, oxide, and sulphate, furnished 
nearly colorless or slightly colored oils with fairly good films, but 
their time of drying was from 36 to 46 hours. Larger amounts of 
these driers than 2 per cent shortened the time of drying, darkened 
the color of the oils, and they did not clarify satisfactorily. 

The acetate and borate are the best of the zinc-salt driers, but 
none of them act catalytically upon the oil as the lead and manga- 
nese driers do, but act meclianically or only while present (like white 
copperas), to throw down some of the mucosities, but do not cast 
them all out, or set up the combination changes necessary to form 
the linoleic compounds required in a good drying oil. The manga- 
nese salts, viz. : the acetate, borate, sulphate, oxalate, and tartrate, 
all gave colorless oils, drying with good firm films in from 20 to 36 
and 40 hours. The citrate and formate gave slightly darker-colored 
oils, drying in about 24 hours with good firm films. The manganese 
borate with quantities varying from 1 to 3 per cent of the oil and with 
temperatures of 220° to 230® F., gave the best colored and drying oils 
obtained in the whole line of experiments. The other manganese 
salts with larger than 1 per cent of driers with temperatures from 
250® to 300° F., gave colored oils of unsatisfactory drying power, 



BOILING LINSEED-OIL. 233 

some of the samples being tarred. No definite conclusion can be 
drawn from Mr. Thorp's experiments as to the relation between the 
quantity of driers dissolved in the oil, and the time of drying of the 
oiL The action of the several classes of driers, as well as the various 
members of them, was erratic and the drying result appeared to be 
governed quite as much by the temperature of the bath, time of 
boiling, and agitation of the oil during boiling, as by the chemical 
employed. One per cent by weight of litharge and the lead driers, 
and two-tenths of 1 per cent of the manganese salts, were all that 
were required to give good bright-colored oils of good drying qualities 
with firm fiJms, and not all of these amounts of driers were taken up 
by the oil, but some were recovered in the residuum. 

The sulphate of zinc boiled with linseed-oil simply removes the 
vegetable and mucilaginous substances that impair its drying power; 
it does not impart any catalytic power to the oil to draw oxygen 
from the air. 

Peroxide of manganese, umber, red lead, and litharge, all dissolve 
in the oil and impart oxygen to it, and act catalytically to take up 
more oxygen from the air to renew what they have lost, and in so 
doing further oxidize the oil. 

Thome and Brin'a Oxygen Process. 

In this process for oxidizing oil, pure, or nearly pure, oxygen 
gas in a finely divided stream is poured through the oil at natural 
temperatures, or moderately heated, if desired. The process occupies 
from 2 to 7 hours, but a small quantity of the usual driers shortens 
the time of oxidation. The color and drying quaUties of the oil 
oxidized by this process are excellent. The consumption of oxygen 
gas varies from 2000 to 4000 cubic feet per ton (250 gallons) of oil, 
according to the degree of oxidation required. 

The principal diflSculty in this process is in generating the supply 
of oxygen gas, which requires a more complicated plant, chain of 
operations, and intelligence on the part of the workmen than that 
connected with the quicker processes of the bimg-hole boil, where a 
bimg-starter, a scoop, and a dish of metallic salts are all that are 
necessary for a manufacturing outfit. 

In the manufacture of corticine (resembling linoleum) the oil is 
oxidized at a high heat, 350® to 500® F., until it begins to thicken 
and get ropy. The result is not only a loss by evaporization, but 



234 BOILING LIN SEED-OIL, 

the oil acquires a peculiar sickish smell that no subsequent treatment 
or step in the manufacture is able to remove. 

The present-day process of boiling oil upon a large scale as prac- 
tised by most of the crushers of linseed is to dissolve 4 pounds of 
lead oxide or litharge, or both, in 5 gallons of well-aged and settled 
linseed-oil at a temperature of about 250° F. for a short time or until 
all of the oxides are absorbed. 

This mixture if allowed to cool will harden into a firm cake of 
gum (linolate of lead). This product while still hot is mixed with 
40 to 50 gallons of Unseed-oil heated to the same degree to coagulate 
the albumin, and the mixture allowed to settle. Five himdred or 
more gaUons of this mixture are made up, and while hot are mixed 
with a large tank (5000 gallons or so) of raw linseed-oil also heated 
to about 200° and thoroughly stirred together. This is commercial 
boiled oil, which varies in character according to the quality of the 
linseed-oil used in any stage of the process, also in the proportion of 
the oxide oil to that in the large tank, 4, 6, 8, or 10 to 1, etc. Lead 
oxide and a small quantity of manganese oxide make a better drying 
oil than the red lead alone. 

Varnish-makers make this liquid drier for use by local painters,, 
who remove 4 or 5 gallons of raw oil from the barrel and replace 
them with the same quantity of the liquid drier, and then roll the 
barrel around or stir up the mixture with a paddle-stick for their 
"bung-hole" boiled oil. 

Some large users of paint oil think that this make of "bung-hole " 
boiled oil is as good as that supplied by the large manufacturers; 
but this is doubtful, as all of the albuminous substances in the bung- 
hole oil are retained unchanged, and they are subject to a future 
decomposition that the 200° of heat in the cooking of the large tank 
of oil coagulates, and they settle out on standing. 

Lead- and manganese-oxide driers made from resin, or resin-oil,, 
are marketed on a large scale at 18 to 20 cents per gallon, while a 
properly made linseed-oil drier cannot be furnished for less than 3 
to 4 times that price. 

Double-boiled oil means that a dorible dose of drier and resin-oil has 
been put through the bung-hole process. The more drier the poorer 
will be the oil product. 

No two manufacturers of boiled oil furnish an oil of the same 
character or quality, owing to their different manner of cooking it 
and the amount and kind of drier used, etc. The name "boiled oil" 



BOIUNG LINSEED^IL. 235 

represents about as uncertain a product as a mixed paint, and is 
simply a trade-name for an extremely varied composition. 

The reputation of the dealer or manufacturer is the best guide. 
Adulterations by the use of resin-oils are to be especially looked for in 
all commercial grades of boiled oil. 

Boiled oil is a misnomer. Linseed-oil never boils; if it did it 
would decompose into a permanent gas. The degree of heat applied 
in the process of so-called boiling by careful manufacturers is only 
that necessary to evaporate some of the water natural to the oil. 
This degree and amount of heat also coagulates the mucilage and 
albuminous substances, so that they are released from the acid oils, 
and, by their greater weight, are deposited as soon as the oil comes to 
a state of rest after the heating process. In oil from ripe linseed, if 
given time to age after crushing, the mucosities and some of the water 
in its composition (about 5 per cent), that is loosely held in com- 
bination, wiU settle and can be filtered out or drawn off, leaving the 
oil bright and clear, and with a minimum amoimt of water to be 
evaporated in the process of dr3ring as a raw oil. Storage in tanks 
for 3 or more months still further improves the oil, especially if the 
tankage is kept at a moderately warm temperature. This fact is 
taken cognizance of by the varnish manufacturers, who require the 
best quality of linseed-oil for their products, and from the better 
prices they receive for their wares, can secure the best qualities of 
oil in the market at a price that the manufacturers of cheap paints 
•cannot compete with. Storage of three or more millions of gallons 
is carried by some of the best varnish and linseed-oil trade dealers. 
The natural result of such selection and storage is, that the oil is well 
aged, clear, and bright, and can be depended upon as a vehicle, unless 
afterward adulterated or abused in its application. 

As stated before, all driers are injurious to linseed-oil, and the 
marked inferiority of trade boiled oil to raw linseed-oil is due to the 
<iriers used, whether liquid or solid, and not wholly to the removal of 
the water and mucosities in the boiling process. The salts and oxides 
of metals that constitute the solid class of driers, that enable the oil 
to dry promptly, are generally added in great excess of the amount 
actually necessary for aiding the natural tendency of the oil to dry. 
That portion of the driers in excess remains in the dried coating, 
and acts as a carrier of oxygen, to attack the pigment; and the subse- 
<}uent failure of the coating is assured. 



CHAPTER XXIV. 

DRYING OF LINSEED-OIL. 

Mulder's experiments in the drying of siccative oils deter- 
mined that there were two periods in the drying of a paint, oil, or 
varnish coating. The first period occurs in the early months, and 
is wholly due to the changes in the drying elements of the oil, and 
while these changes are in progress, the covering is always dry to 
the touch and remains elastic. This period is of longer duration if the 
coating is not exposed to the direct rays of the sun. During the 
period, 100 parts of the oil at ordinary temperatures increase in 
weight to 111 or 112 parts, but when warmed to 170° F. it loses 4 to 5 
parts. In the direct sunlight for all the period of drying, the oil 
gains about 7 parts. 

During the second period the oil becomes hard and firm as a resin- 
ous coating, the change being in the non-drying elements of the 
oil. The increase in weight of the oil is not so great, because the 
breaking-up of the glycerine element has taken place in the first 
period. These total changes amount to about a quart of oil in 1000 
square feet of painted surface". 

The influence of heat in drying a paint or varnish is apparent 
when it is considered that in the ordinary drying of either to a firm, 
hard coating, 21 per cent of oxygen has been absorbed from the 
atmosphere or driers, yet a further exposure to the heat of the sun 
until the coating becomes hard and resinous ensures a loss of 3 to 5 
per cent of this amount. It is to be recommended that this drjring 
and loss be had while the coatings are still elastic, because this loss in 
substance (due to the changes in the non-drying oil) takes place 
while the vehicle is soft and elastic enough to adjust itself to the loss 
in volume. 

Oil or varnish dried in the direct hot ravs of the sun are not as 

durable as when dried in the shade, the effect of the sun being to 

evaporate the volatile elements of the glycerine ethers instead of 

absorbing them into the non-drying or fatty acids of the oil. A 

236 



DRYING OF LINSEED-OIL. 



237 



frost on a drying oil or paint is fatal to its integrity; the coating will 
peel in strips and cannot be restored; but paint dried in clear cold 
weather (not frosty weather, that leaves a sweat in and on the coat- 
ing, as the temperature rises) lasts longer than a sun-dried coating. 

It is probable that the cold-dried coating lasts longer for the reason 
that its change into the hard-soap compound has been delayed, and 
being more elastic than the sun-dried, the bond between the pigment, 
oil, and surface covered is more effectuaUy made. Hence some lead in 
the from of a drier is, therefore, an advantage in an oil compound or 
paint, if the pigment does not contain it. The manganese driera are 
especially imreliable if appUed in cold or unfavorable weather. 

The percentage gain in weight of a cold-drawn raw linseed-oil, 
exposed for drying under different conditions, was as follows: 



Days 
Exposed. 


In Darkness. 


Under 

Unclouded 

Glass. 


Under 
Blue Glass. 


Under 
Yellow Glass. 


Under 
Red Glass. 


Under 
Green Glass. 


10 


.000 


0.126 


0.089 


0.012 


0.009 


0.005 


20 


.001 


0.236 


0.245 


0.041 


0.027 


0.023 


40 


.003 


0.258 


0.376 


0.181 


0.082 


0.139 


60 


.007 


0.298 


0.388 


0.319 


0.178 


0.269 


90 


.013 


0.272 


0.357 


0.417 


0.338 


0.401 


120 


.018 


0.273 


0.360 


0.442 


0.376 


0.438 


150 


.035 


0.300 


0.399 


0.474 


0.441 


0.485 



After 150 days the gain in weight was the greatest in the follow- 
ing order: green, yellow, red, blue, and imcolored. The application 
of a dry heat at temperatures of 150° to 170° F. dries a raw oil from 
30 to 50 times faster than an open-air exposure under the general 
conditions of summer weather. Light, heat, and air are all necessary 
elements to properly dry an oil-paint or varnish coating of any qual- 
ity. The slow-drying oils gain more weight than the quick-drying. 

Raw linseed-oil has a specific gravity of 0.9299 to 0.931, and the 
same oil boiled, that of 9.411. One hundred parts of drying and 
non-dr3dng elements in the raw oil in the process of kettle-boiling 
lose 8 parts of glyceride ether and one part of the carbon and hydrogen 
from the non-drying oil or fatty acids. This is equal to 9 parts in all, 
leaving 90 parts that absorb 21 parts of oxygen from the atmosphere 
when the oil is fully dried, the 100 parts of the raw oil becoming 111 
when dry. 

The presence of the glycerine ether and the changes it eflFects 
between the oleic, stearic, margaric, palmitic, and other acids forming 
the non-dr3ring oil and the dr3ing elements to form the linoleic com- 



238 DRYING OF LINSEEI>-OIL. 

pound, appears to be necessary, as Mulder indicates, who added oxide 
of lead to the oil that formed with the glycerine a hard-lead soap. 
The drying-oil acids were then washed away by ether, the ether was 
drawn off, and the strictly siccative part of the oil set free. When 
this oil was spread and exposed for drying, it gained in weight 8 per 
cent in 3 days, 14 per cent in 7 days, 17i per cent in 30 days, but did 
not become fully dry in many months. 

A putty made from thick glycerine and dry white lead or litharge, 
or both, will harden in from 15 to 45 minutes. 

The conclusions of Mulder and many other experimenters in this 
line, as well as those of practical painters, are: That any metaUic salt 
used as a free or bung-hole drier, or as a heat-combined drier, that is 
so energetic in action as to destroy the glyceride element instead of 
aiding it to effect its change into a drying oil, is an injury. The oil 
can be made to dry by evaporation by adding volatile driers, but it 
does not form a firm reliable coating, Uke that produced by a natural 
resinification. 

The London Journal notes that Lippert, experimenting with oik 
and varnishes in order to observe the eflfect of the absorption of 
oxygen during the process of drying, finds that for boiled linseed-oil 
the less drier used the better, whether the drier be a manganese 
resinate or lead oxide. After its application upon a surface, a coat- 
ing of this kind increases in weight while drying. The higher the 
proportion of drier, the sooner is the maximum of absorption reached, 
and the sooner, also, the coating begins to lose weight again. Con- 
versely, the smaller the proportion of the drier, the larger is the total 
weight of oxygen taken up. After 4 weeks, the varnish containing 
2 per cent of manganese had commenced to soften again, and stuck 
to the palm of the hand. The preparation containing only 0.15 per 
cent of manganese was distinctly the best. With litharge, also, the 
larger the proportion of drier the sooner the varnish attained its 
maximum weight, and the less total oxygen was absorbed. All the 
films were equally hard to the touch when they had reached their 
maximum weight; but after 4 weeks they softened again— this being 
more especially noticeable with the specimens containing much of 
the drier. Additions of litharge above 2^ per cent made no appre- 
ciable difference in the absorbing power; while for practical work 
the proportion of the lead ought evidently to be kept lower than that 
permissible with the manganese salts (see Thorp's Experiments, 
Boiling Oil, Chapter XXIII.). The reason why such coatings soften 



DRYING OF LINSEED-OIL. 239 

again after becoming dry is not yet known. It may be f omid to depend 
on the presence of an excess of drier, or of an unsuitable one. It does 
not prove adulteration with rosin or rosin oil. It may be due to 
oxidation of lead linoleate into a liquid turpentine-like body by pro- 
longed exposure to the air. Science indicates no better way of testing 
a dr3dng varnish than by the finger. 

Prof. Max Pettenkofer removed the non-drying acids from freshly 
dried linseed-oil skins by ether, leaving an elastic caoutchouc-like 
substance, which by degrees hardened and became brittle. On 
further exposure to the air it separated easily into thin flakes; in 
fact, it hardened and cracked Uke the fossil resins or the coniferae 
crude gums. 

These deductions from a long chain of closely and carefully con- 
ducted experiments, not only in the laboratoiy, but in the application 
of oil, paint, and varnish coatings, with all classes of pigments, spread 
over ferric, wood, and mineral surfaces, appear to be ignored in many, 
if not in most, of the present-day compositions of paints and oils. 

The larger amount of mucilage in all unripe or damaged linseed 
and other vegetable oils, when freshly made, or that have been ex- 
tracted by the bisulphide of carbon, the hot or steamed seed pro- 
cesses, deteriorates the quality of the oil, and such oil even if it is to be 
kettle-boiled, is benefited by long standing to allow some of the "mu- 
cosities" or "drops" to settle. The amount of these preliminary 
"foots" is from one to one and one-tenth part in one hundred parts of 
oil. The greener or poorer the class of seeds from which the oil is made 
the greater will be the amount of "foots." 



CHAPTER XXV. 

UNSEED-OIL TESTS FOR ADULTERATIONS. 

There are many methods of testing the purity of linseed-oil. 
The specific gravity test is of small moment, even if not altogether 
imreliable, as the commercial adulterant oils (over one himdred in 
number) vary in specific gravity only seven-hundredths of a degree 
Baum6, and no rehable chemical test has been found that is practical 
for the ordinary analyst. 

One difficulty hes in procuring a chemically pure oil to make the 
comparison with the reactions caused by sulphuric and nitric acids, 
and caustic-soda treatment for the color changes in the sample tested. 
The linseed for such an oil must be picked over with the greatest care, 
selected from fully ripe and full-weight seeds, and pressed in seed bags 
that have never been used before. 

The result of an application to a large number of the most responsi- 
ble oil-seed-pressing firms, was that not one could furnish a sample 
with less than 5 per cent of other seed-oil in it, and most of the best 
commercial brands contained over 12 per cent. 

Approximation chemical tests are numerous, but in all cases a 
sample of known quality of linseed-oil should be used for comparison, 
even if it is not a chemically pure oil. Abbe's and other refractom- 
eters are used to test the purity of linseed-oil, but their use is too 
complicated for any one but an experienced chemist. 

All the fatty oils change color when brought into contact with 
strong sulphuric acid. If a drop be added to 8 or 10 drops of an oil 
placed on a glass plate resting on white paper, the following colors 
are immediately produced. Olive-oil turns a deep yellow, gradually 
becoming green. Sesame-oil, a bright red. Colza-oil, a greenish- 
blue aureola. Poppy-seed oil, a pale yellow with a dingy gray look. 
Hempseed-oil, a distinct emerald green. Linseed-oil turns brown- 
red, changing to a black brown. 

Prof. F. C. Calvert's chemical method of testing oils by acids, 

240 



LINSEED-OIL TESTS FOR ADULTERATIONS. 241 

for the color changes, is given in full in Watts's Diet, of Chem., Vol. 
IV, p. 183. 

Mineral oil or petroleum in any form cannot be added to linseed- 
oil to exceed 5 per cent without affecting its drying; and 10 per 
cent prevents its drying other than as a thin skin impervious to the 
air, and the oil remaining green beneath is liable to blister or peel on 
exposure to sunlight. It does not bond to the pigment or surface 
coated. 

If a tin plate coated with a mixture of a vegetable and a mineral 
oil be viewed at different angles in strong sunHght, the mineral oil 
can be detected by the iridescent or metallic play of color, which 
petroleum imparts to all vegetable oils. So characteristic is this, 
that a little experience will enable a painter to readily detect mineral 
oil when present even in a small quantity. This feature is in some 
degree disguised or palliated by treating the mineral mixed oil with 
caustic lime, chalk, etc., and adding an excess of strong free driers, 
which may suppress the iridescent hues; but if the sample is ignited, 
the marked pimgent odor of burning petroleum vapor is readily 
recognized. 

Cottonseed-oil, a semi-siccative oil, was formerly mixed in large 
quantities with linseed-oil. Its specific gravity, color, taste, and odor 
are almost identical with linseed-oil. It requires a large amount of 
driers either compounded with it by heat or as free driers to enable 
it to dry. Its tendency at the best is to harden and crack the over- 
lying coatings; also to crack in cold weather, from its non-elastic 
condition due to the driers used. Crude cottonseed-oil treated with 
strong ammonia mixed with 3 parts of water, gives an opaque brown 
color. Refined cottonseed-oil, similarly treated, gives a brownish or 
dull opaque yellow. Pure linseed-oil similarly treated for comparison, 
gives a bright, but semi-transparent yellow. Mixtures of both oils 
^ve an intermediatory color, that a little experience will enable a 
painter to determine approximately the amount of the mixture. 

A quick test by cold is to place a sample of the cottonseed-oil, 
mixture on a piece of glass alongside of a known sample of linseed-oil, 
and place the glass in a refrigerator or on a piece of ice. In a short 
time the impure sample will become discemibly thicker than the 
Hnseed-oil. 

Crude cottonseed-oil produces a brown deposit on a piece of 
bright copper foil (to be had from all dealers in chemicals) if left in 
contact with it in a warm place for 3 or 4 days. The price of cotton- 



242 LIN8EED-01L TESTS FOB ADULTERATIONS. 

seed-oil of late years has become so nearly that of linseed-oil that its 
use to adulterate the latter has sensibly diminished. There is, how- 
ever, a combination of the damaged seed oils from both crops, for 
which there is always a demand to furnish a mogrel hnseed-oil at a 
cutrrate price. 

Resin oils are freely used to adulterate linseed-oil, even by firms 
whose business reputations should warrant more honest wares. Resin 
mixes readily with linseed-oil, and whether its grade be Ught or heavy, 
it cannot be detected by its specific gravity. For a quick test of its 
presence, shake up a spoonful of the sample with 5 times the quantity 
of strong alcohol; pour off the alcohol and add to it a clear solution 
of sugar of lead; a cloudy precipitate shows the presence of resin. 

Resin-oil can also be detected by the remarkably nauseous after- 
taste produced by it when touched by the tongue. The odor of the 
oil is not recognizable in the sample unless ignited, when it becomes 
decidedly different in odor from ignited linseed-oil. If the cork of 
the sample bottle squeaks when it is twisted around in the neck of the 
bottle, no further test is necessary to denote its presence. 

Linseed-oil adulterated with resin-oil of any grade is readily de- 
tected by passing a current of chlorine gas through the sample oil, 
which is rapidly blackened if any appreciable amount of resin is present. 

Resin-oil is especially to be looked for in boiled oil. It never 
hardens completely, and makes the coating "tacky." If any consid- 
erable amount of resin is present in a linseed-oil that has received an 
excess of driers to harden the coating, the coating will be brittle and 
crumble easily on a short exposure to the atmosphere, particularly 
in summer weather, or by exposure to heat. 

Menhaden and porgy fish-oils are used freely to adulterate linseed- 
oil, especially in many of the mixed paints and pastes. The price 
of these oils is only about one-half that of a poor linsccd-oil. Fish- 
oil in the twentieth century, used as an adulterant of linseed-oil, com- 
prises almost anything from a whale to a mossbunker, with the oil 
from dead animals frequently added to help out the abomination. 
The fish-oils dry slowly, but surely, if fortified by strong, free driers. 
Fish-oils used for a tin-roofing paint will stick longer than a linseed-oil 
paint, as they do not dry so hard. They are more affected by cold 
than linseed-oil, and whatever paint coating is spread over one with a 
fish-oil vehicle will probably peel in a short time. 

Crude menhaden-oil when cold has but little odor, and in color 
closely resembles linseed-oil. By strongly heating a sample the fishy 



L1NSEEIU)JL TESTS FOR ADULTERATIONS. 243 

odor is developed. Skilfully mixed, the odor is hard to detect, even 
when moderately heated. 

Place a sample of a known quality of linseed-oil and one of the 
oils to be tested, in separate test-tubes, coik, and then heat together 
in a hot water-bath; if the suspected oil gives off an odor of acrolein 
(oxidized glyceride) similar to that of the smoldering wick of a tallow 
candle, fish-oil of some grade and amount is present. 

One of the most reliable tests of the purity of linseed-oil, and one 
that does not have to be felt for as in the preceding tests, and is equally 
available, is, viz. : Add to 100 parts of the oil by weight, one-half of one 
per cent by weight each, of litliargeand red lead well stirred together. 
Heat in any convenient vessel, and in any manner, until an immersed 
thermometer reaches 480° to 500° F. A feather from a feather duster 
or chicken's wing will answer instead of a thermometer. If the 
feather when dipped in the hot oil curls up with a crackling sound, 
it indicates an approximate temperature. A small current of air 
from a bellows or other source should be blown through the oil as it 
is being gradually heated. A small glass tube and a piece of drug- 
gist's rubber tubing are readily available for this part of the appa- 
ratus. Small samples are taken out from time to time and cooked 
on an iron plate. As soon as the samples appear stringy when cold, 
allow the oil to cool, stirring it constantly during the cooling. If the 
oil is solid and firm when cold the sample is of good quality. A poor 
oil will be sticky and more or less fluid, and of bad odor. 

Animal oils can be detected by their odor when the sample is 
heated, also by the addition of nitric or sulphuric acid, either of which 
gives an intense-red color to fish-oils. The adulteration of mineral 
oils is readily discovered by the process of saponification, when these 
substances rise to the surface. The saponification number, or the 
number of milligrams of K.HO required to saponify linseed-oil, is 
190.2 to 192.7. This number for many adulterated oils is as low as 
180. The iodine number of linseed-oil is 158.7 to 159.78; that of 
fatty acids, 159.85. 

It is frequently necessary to clarify linseed-oil. The following are 
a few of the methods : 

Heat the oil slowly up to 300° C. (570° F.) either by itself or with 
the addition of from 1 to 5 parts of either caustic lime, carbonate of 
lime, calcined magnesia, or carbonate of magnesia, and keep the oil 
at the above temperature for 2 hours, and then allow it to cool uncov- 
ered and undisturbed. Transfer it to a settling-tank to deposit and 



244 CLARIFYING LINSEEDjOIL, 

clarify. When clear transfer it to another vessel. Clarified oil should 
not be kept in contact with the deposit or "drops." 

Mulder recommends the clearing of turbid linseed-oil by washing 
it with its own volume of warm water containing some common salt. 

Sulphuric acid is used for clarifying linseed-oil, particularly adul- 
terations of it. Its action is to carbonize the fibrine elements of the 
fish-oils and the mucilaginous substances in the vegetable oils, and 
to deposit them in the so-called "foots" or "drops." Its action is 
injurious to linseed-oil in general, for it removes by carbonization 
a part of the fatty acids, or non-drying elements, and all of the 
glyceride ethers, the latter being essential to the changes necessary 
to form a firm, hard coat from the oil when dry. Acid-treated oils are 
liable to dry on the exterior only, and never become hard or firm. 
Acid-treated oils require long, careful, and repeated washings with 
warm water in the form of an air-blown spray through the body of the 
oil in a deep tank to eliminate the acid, which is seldom thoroughly 
done. The acid, besides delaying the drying, will attack the metallic- 
base pigments afterward associated with it in the paint. 

Graphite and carbon pigments are less affected by sulphuric acid 
in the oil than any other class of pigments. 

Graphite paint-skins detached from the metallic plates on which 
they had been spread and dried, when immersed in 5 per cent solu- 
tions of sulphuric acid, lost in weight from 1.5 to 1.7 per cent, but 
were not affected in lustre, strength, or elasticity. 

This result indicates their superior qualities for heavy coatings 
for roofing paint, and in other locations where any sulphur element 
can reach them, whether from the vehicle or the atmosphere. 

The results of Mulder's experiments with sulphuric-acid-treated 
oils to ascertain their qualities were that they did not begin to dry 
materially under 8 days. At the end of 3 months the samples had 
gained 15 per cent in weight, but lost 3 per cent of this when heated 
to 150° F. for a short time. Also a strong heat from the sun for a 
number of summer dajrs produced the same effect. 

Pure linseed-oil, not treated, gained 10 per cent in the same 
period and lost 2.5 to 3 per cent on heating it. The acid did not 
affect the drying elements in the oil, only the non-drying ones, as 
noted before. 

The colors of the acid-treated oils were not materially aflFected by 
the process; generally, they were brighter for the treatment. 

Sulphuric-acid-treated oils being naturally of a fatty or non-dry- 



LINSEED-OIL SULPHURIC-ACID OILS, 245 

ing character, if ground in a hot mill will develop this feature more 
fully. 

Sulphuric acid in the oil or pigment appears to take kindly to 
wooden surfaces as a priming coat; at least the disintegrating effect 
of the acid is not so marked as upon metal. All sulphuric-acid-treated 
oils have a tendency to increase the galvanic action in aU paints made 
from them, that are spread on wooden surfaces. When used on 
metaUic bodies electrolysis is increased. 

The foUowing table * gives the weight in grains of sulphur in a 
gallon, of a number of the oils when burned by means of a wick floating 
in the oil, and condensing in a sulphur apparatus the vapors of com- 
bustion, the same way as sulphur is determined in coal-gas: 

Name of OU. ^"T« offi'" 

Linseed-^il (La Plata) trace. 

CottoDseed-^il : . trace. 

Olive-oil none. 

Groundnut-oil none. 

Sperm-oil, ordinary 2.3 

" " bottle-nose 3.1 

Cocoanut-oil 3.7 

Neatsf oot-oil 4.7 

Cod-oil 5.8 

Rape-oil (Jamba) 11.3 

" " pure brown 14 . 2 

" " ordinary brown 17.4 

" " brown, refined with sulphuric acid 16 .8 

" " " " " fullers' earth 10.0 

" " " (Ravision's) 19.1 

Russian nuneral oil, crude, 0.908 20.5 

" " " burning 10.3 

American mineral oil, water-white 8.1 

" " " burning 16.3 

" " " safety 14.0 

Scotch mineral oil, used for making gas 49.8 

Water in Linseed-oU. 

Pure raw Unseed-oil contains over 5 per cent of combined water, 
and the commercial or poorer grades of the oil frequently contain 10 
per cent. Twenty per cent of water can be stirred into linseed-oil by 
a painter's paddle, and over 10 per cent more, if the mixture is run 
through A mixing-mill, either with or without a pigment. 

* Chmnical News, also Scientific A merican Supplement, July 20, 1895. Experi* 
ments by Wm. Fox and D. G. Riddick. 



246 WATER IN LINSEELU)IL, 

Mixed white-lead paint (pure or adulterated) wiU form an emul- 
gion with its own bulk of water if run through the mill, and the water 
does not separate to any great degree, unless it stands for many weeks, 
and then being at the bottom of the can or package, escapes notice. 

The covering or spreading power as well as the coloring power of 
such mixtures are of the poorest character; at least an extra coat and 
often two are necessary to present any sort of a creditable appearance. 
They dry solely by evaporation through the outer skin of the paint, 
leaving it porous, and where moisture can escape, the same element 
containing other atmospheric gases can enter and they are more de- 
structive than the moisture they replace. All water-oil mixtures dry 
flat and lifeless. The use of alkalies and strong metallic-salt driers 
to form a better emulsion does not materially change their forced 
mechanical association, and if they cannot evaporate and escape out- 
wardly, they go inward and condense on the covered surface, form 
blisters, and peeling results. 

It is almost impossible to spread a water-oil paint in the cold 
without heating it. If spread and not dry, a cold night, not even 
approximating a frost, wiU ensure a blistering and peeling the next 
day. Brushing the coating over with turpentine or benzine will not 
prevent or correct this action, which will take place regardless of the 
nature of the pigment. A good Unseed-oil paint spread on a cold day 
(not freezing weather) will "take" and dry if a little extra eflfort on 
the painter's part is made to brush it out well and by using a little more 
turpentine for the drier than that used in warmer weather. But all 
painting for external exposures should be suspended in cold weather, 
especially on all ferric structures, unless under cover and in a warm 
room, where the painted surface should remain until the paint has at 
least " set " firmly, or until it has dried enough to bear handling 
without feeKng "tacky." 

Many paint chemists deem that 2 per cent of wat«r added to 
linseed-oil in excess of that natural to it, whether made from ripe or 
unripe linseed, is not detrimental to a paint. To enable the oil made 
from unripe seed to carry the extra water, it is rendered slightly alka- 
line» generally by adding lime-water, which forms with the oil vehicle 
a calcium soap that thickens the paint, so that it never settles hard, 
and is easily stirred up, consequently, does not dry hard. 

A number of tests for free water in linseed-oil are: Heating the oil 
to 212° F. for a short time, and note the amount lost by evaporation. 
Filtration of the oil after heating and the addition of dehydrated copper 



WATER IN LINSEED-OIL. 247 

sulphate, which turns bright blue when added to the oil. A strip of 
gelatine muneised in the paint for 10 or more hours will absorb the 
free water and sweU up. Cool the paint or oil for a few hours in a 
refrigerator, and note the difference in its spreading. Heat a piece of 
iron to a bright cherry-red and plunge it into the oil or paint. If 
there is much snapping, it indicates the presence of free water in the 
mixture. In ordinarily pure oil or good paint, a thick, heavy smoke 
without explosion or snapping will follow the withdrawal of the glow- 
ing test-iron. 



CHAPTER XXVI. 

SUBSTITUTES FOR LINSEED-OIL. 

Many so-caUed substitutes for linseed-oil have been presented to 
the pubUc in the past, and at present they are numbered by the hun- 
dreds in the Patent-oflSce records, and outside in the formulae of the 
compounders of paints. All substitute oils are as uncertain and indefi- 
nite in character as the pigments assembled with them. 

Generally, a low grade of linseed-oil is the base for the vehicle, to 
which one or more of the score of flax-dodders or buffums, resin, fish, 
mineral, india-rubber, and soap-fat oils are added. These are mixed 
with benzine for the volatile, also with manganese or other strong 
metallic-salt driers, put through the bung-hole. No heat is employed 
in their manufacture, and some of them are dangerous to sell, or spread 
even when cold. 

Purchasing agents and master painters (except in a few cases) 
have not the laboratory, chemical apparatus, technical knowledge, 
or time to analyze them to detect the fraud. The result of their use 
comes with the lapse of a very short time, when the scraper, burning 
torch, and a new coating are the only remedies for the evils of crazing, 
peeling, or decomposition due to galvanic action. No amount of 
sophistry can change the fact that the use of so-called substitute 
oils has in nearly every case been detrimental to the paint and the 
covered surface. 

Probably 80 per cent of all the oils, paints, and varnishes manu- 
factured in the world is applied to structures of minor importance, 
which are destroyed by causes other than corrosion. These coatings 
are quite as much for looks as for physical condition, but the other 
20 per cent is used on the most important and costly engineering 
structures of our time. These require protection from corrosion 
from the hour the materials leave the rolling-mill, forge, and foundr}'-, 
imtil they are in the finished structure, and need more then than 
during construction. 

248 



SUBSTITUTES FOB LINSEED-OIL. 349 

Among the recent substitute or paint oils (dating from 1895) is 
the substance called "Lucol," for which extraordinary claims are 
made, viz.: 

"Lucol" * is a paint oil, a synthetical compound, made by a secret 
process from materials that necessarily are a part of the manufac- 
turer's secret." 

What it is as a chemical compound, or as a manufactured paint 
oil, concerns the proprietors of it. What its claims are for superiority 
over linseed-oil, concerns the users. As set forth by its manufac- 
turers, its characteristic features in comparison with linseed-oil are: 

" 1. Lucol is more durable than linseed-oil, which dries by absorp- 
tion and oxidation and only to a small degree by evaporation. The 
reverse is the case with lucol, which dries principally by evaporation, 
hence the condition of the atmosphere to which it is exposed while 
drying must be considered. 

"2. Lucol sets quicker than linseed-oil, which is the result of 
evaporation instead of the oxidation of its elements; the final drying 
to a bone-hard condition requires many months. The retention of 
its elasticity no doubt accounts for its durabiUty. 

"3. Lucol sets sooner and dries quicker than linseed-oil, hence 
is less adhesive for dust, and does not wash off if rained on, as is the 
case with other vehicles. 

"4. Lucol gives a purer white with white lead than linseed-oil. 

"5. Lucol preserves the original tint of the pigment longer. More 
lucol is used in a coating than when linseed-oil is used. The gloss is 
less at first than with hnseed-oil, but it is soon ahead in this respect. 
The oil is the life of the paint. 

"6. Lucol can be flatted with a much smaller proportion of tur- 
pentine than with linseed-oil." 

Other advantages are set forth, but all are more adaptable for the 
use of lucol on surfaces other than ferric, and have been answered 
elsewhere in this work. 

The Lucol Company says: "TFe extract the olein, which is 
carefully refined by a special and partly secret process, and by 
the use of chemicals it is converted into a brilliant, transparent, 
lemon-colored oil. It contains no vegetable, mineral, or jish oil, resin 
or resinroil, varnish-gums, benzine, or other powerful solvent driers. 
Ignited it gives off an odor similar to that of burning india-rubber, 

* " Lucol." Excerpts from Painting and Decorating Journal, New York, 
February, 1895. 



250 LUCOL AS A PAINT OIL. OLEIN. 

entirely different from the odor of linseed-oil in combustion. It has 
an unpleasant odor while being spread and in drying, wholly unlike 
linseed-oil, and should be flowed on similar to a varnish, instead of being 
brushed out like a linseed-oil coating. 

"Lucol, in the form of a paint, resists alkaline substances, sea- 
air, sea-water, and covers galvanized-iron surfaces without peeling. 
Lucol weighs 7J poimds per gallon, and is placed on the market on its 
merits as a synthetical manufactured oil, wholly unlike any other sub- 
stitute compound heretofore offered as a paint-oil." 

How well are the above claims founded? "We extract the olein," 
etc. This at once destroys the claim that lucol is a S3aithetical com- 
pound. It is only an oil made with a vegetable, an animal, or a 
fatty acid base. 

All of the solid fats and oils are derived from two sources. The 
marine animal oils are obtained from the cold-blooded fish, like the 
cod, menhaden, etc., and the hot-blooded, like the seal, sperm, and 
right whale, etc. The terrestrial animal oils are lard, neat's-foot, horse- 
bone, tallow (oleic acid), etc. Both classes may be considered as 
mixed glycerides of oleic acid (Ci^K^fii)} stearic acid (CigHjeO^), and 
palmitic or benic acid (QoH 3202)5 the first preponderating in the 
oils and the two last, especially the stea.ric (steaune), in the fats. 

Oleic acid has a specific gravity of 0.808 at 65° F. and is the liquid 
acid obtained by the saponification of non-drying oils and liquid fats, 
which contain a different glyceride than the drying oils. The propor- 
tion of olein differs acco ding to the nature of the fat f om w^hich it 
is obtained. 

Chevreul prepared it by boiling human fat, lard, goose-fat, beef, 
and mutton suet, filtering the solution and allowing it to stand for 24 
hours, then concentrating it a little by evaporation, adding water to 
separate the olein, and separating the liquid from the solid' matter by 
pressure. Olein thus obtained does not solidify at 32° F. 

Olein is also prepared from olive-oil and other glycerides contain- 
ing it by pouring upon the fat a cold strong solution of caustic soda, 
which saponifies the solid fats but not the olein. It is also obtained 
from olive and almond-oils by treating them with cold alcohol and 
evaporating the solution. 

Pure olein is a colorless oil void of taste and smell, insoluble in 
water, very soluble in absolute alcohol or ether. Specific gravity, 
0.90 to 0.92 ; bums with a bright flame, oxidizes in the open air, 3nelding 
the same products as oleic acid. Crude, or carelessly prepared, the 



SUBSTITUTES FOR LINSEEJ>OlL. OLEIN. 251 

olein will have an odor distinctive of the class of fats from which it 
is obtained. 

The marine oils all have the repulsive fishy odor in various degrees, 
sperm-oil being the hardest to locate. The terrestrial animal oils 
have the peculiar sourish odor of cooking fats. The vegetable oils 
have a sweetish odor. A Uttle practice with a heated sample will 
enable the most of them to be recognized. 

Olein is also extracted from the organic acids in soap-stock or the 
fats left in the by-products in the refining of cottonseed-oil. The 
fatty acids in the ** foots" are distilled with superheated steam; when 
the distiUate cools and soHdifies, the olein is extracted by pressure. 
The process is analagous to the production of commercial cottonseed- 
oil and lard stearins used in the preparation of butterine, lard surro- 
gates, and candles. There are about 250,000 gallons of olein available 
in the cotton crop of the United States, if aU the foots were used for the 
extraction of olein and none used for the manufacture of cheap soaps. 

There is no amount of animal oils or fatty refuse available for 
manufacturing into olein that can in any material way affect the 
broad field covered by linseed-oil as a vehicle for paints. 

Chemistry has not arrived at that stage of development where the 
assembling of the chemical elements of fatty substances in their known 
proportions will produce an oil or fat. All such substances must have 
a natural base for the foundation for the chain of operations and 
reactions necessary to change their nature. 

Lucol, as a paint vehicle, therefore, is not a synthetical com- 
pound, but a manufactured paint oil. Its endorsement by master 
painters when used on passenger-cars or other works which are cov- 
ered by coatings of fossil-gum varnish, or upon ferric bodies which 
are thoroughly covered with rust, is no evidence of its resistance to 
corrosion. A few successful applications of it under favorable condi- 
tions wiU not counterbalance the failures. Generally, no record of 
the failures of substitutes for linseed-oil is available for the public, who 
are as much interested in knowing what not to use, as what vehicle 
is the best for a paint. 

A siccative oil of peculiar properties has lately been introduced 
from China into England and the United States. Its comparison 
with linseed-oil for paint and varnish coatings is as follows: 

Chinese wood-oil * has thus far been employed for the manufacture 

* " Uses of Chinese Wood-oil in the Manufacture of Paints and Varnishes." 
TTaosAated from the Fdrben Zeitungf by the Scientific American, January, 1895. 



252 SUBSTITUTES FOR LINSEED-OIL. CHINESE WOOD-OIL, 

of lacquers, varnishes, and paints, on account of its peculiar quality 
of drying thoroughly in about 6 to 8 hours. Pigments ground in the 
oil furnish excellent paints, that do not remain soft and sticky below 
the surface, like coatings prepared from linseed-oil. 

Chinese wood-oil is favorably employed as a floor oil or paint on 
account of its hardness; also in the manufacture of an oilcloth-like 
goods, which, when dried in hot air, excel the ordinary oilcloth or 
water-proof products by their extraordinary elasticity. The odor of 
the oil is very peculiar, resembling lard, and remains in the coating 
for months, and even for years. This lard odor remains in the lacquers 
made from the oil; hence for that use, also for floor and other interior 
uses, it is necessary to remove it. Disguising the smell by the use of a 
volatile oil does not answer the purpose, because the odor reappears 
after the evaporation of it. Among the remedies resorted to are: 
Agitation with a dilute solution of permanganate of potassiiun; a 
filtered solution of chloride of hme, filtration through animal char- 
coal; mixing with potato flour, also storing it for a long time after 
filtration, after the process of Bang and Ruffin. It is also possible 
to obtain a tolerable freedom from the odor by the use of a blower 
passing air heated to not exceeding 50° C. through the oil, for 6 to 8 
hours, when it loses perceptibly in odor and can be used for lacquers 
or floor-work. For outside exposures it is not necessary to attach 
much value to the deodorization. 

Wood-oil in its raw state dries opaque, probably due to the presence 
of mucilage and albumin. In this state the oil becomes waxy at low 
temperatures, and organic salts analogous to the stearates settle out. 

Wood-oil is boiled for a short time in a Hke manner to linseed-oil 
with a small percentage of red lead or litharge, else it will always remain 
opaque. This for paints is inunaterial, but boiling is necessary to 
give a greater drying quality. 

In boiling the oil, whether with lead or manganic compounds, a 
temperature of 200° C. must not be exceeded, otherwise, in the use of 
borate of manganese, a thickening ensues, followed in a short time by 
complete gelatination and waste of the oil. Heats approximating 
160° C. and in any case not over 180° C. should only be used, and for 
but a short time, when the oil should be removed from the fire and 
the siccatives stirred in. This imparts to the oil a drying quality and 
obviates gelatination. Pigments can be ground in the oil as usual 
with the use of linseed-oil. Compositions of linseed-oil and wood-oil 
work well together, being especially adapted for exterior varnishes 



CHINESE AND JAPANESE WOOD-OILS, 263 

on account of the hardness, solidity and quick drying they receive 
from the wood-oil, and the elasticity from the linseed-oil. 

The important drying quality renders wood-oil useful in the 
manufacture of fatty lacquers. It cannot be employed for spirit 
lacquers, as it is insoluble in alcohol. 

The wood-oil of China and Japan is obtained from the seeds of 
the tung-oil or varnish tree {AleurUes cordata, Elaeococca vemida). 
Another variety, the AleurUa trUoba, furnishes an oil of less drying 
power, and is used to adulterate the oil from the former. 

About 266,700,000 pounds of the oil are annually shipped from 
Hankow to other parts of China, and for export. The Canton wood- 
oil is said to be better and purer than the Hankow, and is about 10 
per cent higher in price. The cost of these oils in England, where 
their use is firmly established, is from 4 to 4^ cents per pound. 

De Negri and Sburlati report that the fruit of the tree from 
which the oil is extracted contains 53.35 per cent of oil, 42 per cent 
being recoverable. 

The cold-pressed oil is of a pale-yellow color, is tasteless, and has a 
smell like castor-oil. The hot-pressed oil is a medium brown in color, 
with a taste and smell like hog fat. 

The drying power of the oil is superior to that of linseed-oil, the 
cold-pressed drying better than the hot-pressed. 

Its specific gravity is 0.936 to 0.941. Saponification value, 156.6- 
172. Iodine value, 159-161 (de Negri and Sburlati). The oil is 
soluble in cold absolute alcohol and mixes readily with linseed-oil. 
When heated with Htharge it turns darker in color and evolves a 
slight smell of acrolein. When thinly spread on glass in a closed room 
it dries to a whitish film resembling milky or frosted glass. A heavier 
coating exposed in the open air dries in about 6 hours. The oil after 
heating or boihng by itself develops the whitish film, but when boiled 
with litharge is as clear and bright as any oil varnish. 

Very thick layers of the dried oil can be scraped off as a tough mass 
quite uniform throughout. It has an exceptionally small adherence 
to glass. 

The balsam known as wood-oil or gurjan balsam, from the Dip- 
terocarpua turbinatiLSy Gaertn, should not be confounded with the tung- 
oils. It is frequently adulterated by them. It is also a natural 
varnish. 



254 NATURAL VARNISHES OR PAINT OILS. 



Euphorbium. 

This substance is in its experimental stage in the United States 
as a vehicle for anti-corrosive and anti-fouling paints. Attention 
was directed to its anti-corrosive qualities as a natural varnish and 
its probable utility as a vehicle for paints about the year 1870, from 
the discovery that the axes, machetes, and other tools used to cut 
down the thickets of the euphorbia spurge, to clear the way for a 
survejdng expedition in Natal, became coated with a strong glutinous 
juice that adhered so firmly to them as to be with difficulty removed. 
The tools coated with it did not rust in fresh water, and bilge-water 
had but little effect upon it. When articles were coated with it and 
immersed in the sea, no barnacles or marine fife would adhere to it. 
Its effect upon insect fife appeared to be equaUy repulsive, and timber 
coated with it resisted the ravages of the Teredo navcdis. It resists 
heat and cold better than finseed-oil and varnish vehicles, while 
ammoniacal and chemical vapors do not cause blistering, scafing, or 
other injurious action. 

Euphorbium juice has a strong affinity for iron and steel, and 
when applied in its crude state as it exudes from the shrub, has no 
injurious effect upon metals, wood, or other substances used for engi- 
neering or common building purposes. When prepared for a paint, 
the juice undergoes several special processes and becomes a clear 
gummy vehicle of a medium-brown color, that receives the usual 
color pigments much as finseed-oil does; retaining, however, its own 
protective properties unimpaired. 

Euphorbium, prepared for a vehicle, appears to maintain its 
properties in all climates, and does not apparently deteriorate with age. 
It is perfectly elastic, tenacious, and when dry, can be drawn out to a 
thin thread. It adheres firmly to polished steel, tin plate, zinc coat- 
ings, sheet lead, and spelter. Earth acids appear to have fittle effect 
upon it, as pipes coated with it and buried for a number of years 
show fittle injury. 

Euphorbium juice has a bitter, biting taste, is very acrid and 
irritating to human flesh, corroding and ulcerating the body wher- 
ever it is appfied. The sores resemble those from nitric acid, and are 
hard to heal. In this and nearly all other respects it resembles the 
crude juice gathered from the Rhus vemicifera, called by the Japanese 
vrushz-nakif the native lacquer-tree of Japan. 



NATURAL VARNISHES. EUPHORBIUM. 265 

The euphorbium of commerce is imported in casks, and is a gunmiy, 
resinous substance in the form of drops of an irregular size resembling 
gum arable. The drops contain vegetable matter — ^twigs, flowers, 
thorns, etc., that collect on the gum as it exudes from the tree and 
are dried in; though many of the tears are hollow and without refuse 
in them. The natural color of the tears is a cloudy pale yellow exter- 
nally, but of a lighter color internally. The tears break easily in the 
fingers, but are difficult to pulverize. The principal part of the process 
to prepare the crude gum for use as a varnish or a vehicle for paint is 
to free it from the vegetable matter. It is partially dissolved by 
water and almost entirely by alcohol, ether, and oil of turpentine. Its 
composition by analysis is, viz.: 

A resin soluble in ether 26 .95 per cent. 

" " insoluble" " 14.25" 

Euphorbin, the peculiar principle 34 . 60 " 

Caoutchouc 1 . 10 " 

Malic acids 1 .50 " 

Gum salts 20 .40 " 

Ammonia soluble matters 1 . 20 " 



100.00 " 



It has no acid reaction but an extremely burning taste. 

The intense acridity is due to the resin, soluble in the ether, which 
melts at 107.6** to 109.4° F. The resin insoluble in ether melts between 
246.4° and 248° F. Euphorbin is a crystallizable substance, fusing 
at 154.4° F., and soluble in ether, benzine, etc,, but not in hot water. 

Euphorbium dries readily without the use of metallic salts or 
solvent driers. Ox-gall and other kindred substances appear to be 
the best driers for it, when any are required. The crude resin is the 
product of the Euphorbiaceoe: the genus is numerous. There are about 
600 species, many of which are natives of nearly every country in 
the temperate zones, and are commonly known as spurgeworts. 

The euphorbium spurge, or E. resinifera, is a shrubby and herba- 
ceous succulent, frequently covered with thorns and having stalks from 
3 to 6 and sometimes 10 feet in height, and grows in almost impene- 
trable thickets in the hot interior deserts of Morocco and other hot 
climates. Euphorbium is obtained from the incisions made on the 
plant; the corrosive milky liquid hardens on the stems in resinous 
drops or tears, or like spruce-gum deposits, and is collected in various 
ways. 

The commercial supply comes principally from the southern prov- 



256 NATURAL VARNISHES. EUPHORBIA. 

inces of Morocco, in the districts of Aitaitab and Juteefa, at the foot 
of the lower range of the Atlas Mountains. 

The euphorbium gum from Natal is considered to be inferior to 
the Morocco product. North Africa is capable of producing euphor- 
bium in sufficient amounts to supply almost any demand for it. 

The gum is called in Arabia "Darkmows," and is known in the 
Eastern markets as "Farfium." 

In India there are 116 species of the AnacardiacecB referred to 23 
genera, in addition to the E. resinifera and some other varieties found 
in Arabia. 

The E. dracuncuUndes in India (jy-chee) yields 25 per cait of 
a clear oil of a yellowish or greenish-yellow color from the dry 
husked seed. The oil is more limpid than linseed-oil, does not 
become ropy from age, and is used for a burning and drying oil. 

The E. lathyris is raised on the edges of the fields in France, 
Germany, and Switzerland. It contains 40 per cent of a fluid oil of a 
siccative nature. 

The E. neriifolia grows wild in Burma, Baluchistan, the Malay 
Islands, etc. It 3delds a gum of a gutta-percha nature on boiling 
the stems and twigs of the shrub. 

The E. Royleana is a large fleshy shrub common on the dry rocky 
hillsides of the Himalayas, growing at an elevation of 6000 feet. The 
sap of this plant yields a superior gutta-percha. 

The Pisticia Lentiscus yields the resin mastic. 

The MelanoTshcea usitalissima yields the black varnish of Burma. 

The Indian Holigama longifolia also yields a varnish. 

The Indian Odina Wodier is covered with its brown gum, which 
streaks down the stem and ultimately turns black. 

The E. pulcherrima is an ornamental shrub grown in Mexico that 
3delds a milky sap which hardens into a black gum, and can be boiled 
down to a sort of gutta-percha. Guatemala and other countries near 
the torrid zone also have a large number of trees that furnish the 
natural varnishes, though no attempt has been made to bring them 
into commercial importance. 

The P. terehinthus is a tree growing along the shores of the Medi- 
terranean Sea. It furnishes the cjrprus turpentine. 

The Japanese lacquer-tree, or the urushi-naki, is known in China 
as the Tsi-chou. It belongs to the botanical order of Anacardiacece, 
to which also belongs the Rhus vemicifera, a tree with very long, glossy 
leaves resembling those of the ordinary sumach, poison-oak, dog- 



NATURAL VARNISHES, JAPANESE LACQUER. 257 

wood, ivy, etc. The American dogwood was formerly thought to 
be of the same species, but is now placed in another of the same order. 

In Japan * the lacquer-tree grows to the height of about 30 feet, 
and at the age of 40 years is about 40 inches in diameter. It reaches 
its greatest perfection in the yield and quality of the lac or varnish 
at the age of 18 years. 

The crude lac, called ki-urushi, is collected at any time between 
the months of April and October by making a number of horizontal 
incisions in the bark of the tree in a manner similar to the "boxing" 
practised to gather the sap of the long-leaf pine- or turpentine-tree. 
The tree is hacked in this manner for from 60 to 80 days, or until it 
dies, when it is cut down, the bark and sap-wood removed and steeped 
in hot water, which extracts the last remnant of the lac, about half a 
pint, which forms the poorest quality of lac, known as "roiro-wrus/iiy 
or black varnish. The tree seldom survives the first season's hack- 
ing, at whatever age it is done. 

The varnish-tree is probably native to CSiina, but it is also found 
native in Japan, and is cultivated all over Nippon and in several 
districts of Kiushia and Shikoku, and there are extensive planta- 
tions in the valley of Tadamigawa and Northern Echigo. A tem- 
perate climate seems to best suit the growth of the tree, as it reaches 
its greatest perfection on the main island north of latitude 36®. It 
is cultivated in Northern Hondo, between 37° and 39®. It may be 
of interest in considering the question of habitat to note that the 
Rhus vemicifera, mentioned by Mr. J. J. Rein, are growing in Ger- 
many at Frankfort-on-Main and at Strasburg. They endured the 
hard winter of 1879-80, when the temperature reached 27® C. 
In Japan the lowest temperature in Northern Nonshiu is — 12® C. 

The lac is purified by straining it through cotton cloth, evapo- 
rating the water by exposure to the sun or by a gentle heat. Some- 
times water is added to the crude lac, and they are ground together 
on a paint slab, and then the water is evaporated. Various coloring 
matters are added to the purified lac by grinding, as is usual in the 
manufacture of oil paints. Black lacquer other than that furnished 
from the last run of sap is produced by the addition of some salt of 
iron. 

♦ Excerpts from "Japanese Lacquer and the Vamish-tree that produces it." 
A communication to the author from the Bureau of Forestry, United States 
Department of Agriculture. By Geo. B. Ludworth, Chief of Division of Forestry 
Investigation, June, 1902. 



258 NATURAL VARNISHES. JAPANESE LACQUER. 

Whenever driers are required, a little oil of tea is iised, also the 
gall from pigs and oxen, to give body to the lac. The purest lac is 
from the first run of the sap after tapping. It is called nashyirurushi, 
and is bleached in shallow vessels laid in the simlight. The other prin- 
cipal grades are the henki-urushi, the unbleached jeshimi-wruahi, and 
the roircMirushi, or black varnish. 

There are about 20 different grades and qualities of these lacquers 
in the Japanese market, of which the above are the principal ones. 
They vary in color from a light brown to a deep jetty black. 

Lacquer is thinned only by heating. The addition of water 
thickens the lac into a jelly. Lacquered objects are always hardened 
in a himiid atmosphere, such as a room with wet cloths himg on the 
walls, or containing a spray or vessels of water. 

All varieties of varnish-trees are propagated by the seeds and 
cuttings. The seeds are gathered in October and sown early in the 
spring, make 10 to 12 inches' growth the first season, and in 10 
years are 9 to 10 feet high and from 2 to 3 inches in diameter. In a 
favorable soil the annual height-growth during the first 6 years is from 
20 to 30 inches, and diminishes afterward to from 10 to 20 inches. 
Plants from root cuttings grow more rapidly than seedlings, but the 
latter make hardier and longer-lived trees. The trees after the first 
5 or 6 years require very little care, and are generally tapped at any 
period after the tenth year of their growth, though sometimes it is 
done when only 4 or 5 years old. 

The climate and soil of at least one-third of the United States 
are as favorable for the growth of lacquer-trees as those of Japan or 
China. Their cultivation requires no more care than that given 
to the sugar-maple or the Eucalyptus. Specimens of the trees are 
growing in the grounds of the Department of Agriculture at Wash- 
ington, D. C. 

Plants which are largely cultivated in Europe have been confused 
with the Japanese Rhus vemicifera. They are, however, a differ- 
ent variety of the tree. The Ailavihus glandvlosa Desf., in France 
called Vemis de Japan, is also of the varnish-bearing species. 

The poison - sumach, Rhus vemeala, common in the Eastern 
United States, >nelds a sap that furnishes a black, lustrous, durable 
vamish, ver>'^ similar to that derived from the Japanese tree. 

Other trees that belong to the same botanical order {Aruicardiacece) 
that jrield natural varnishes have been referred to in the article 



NATURAL VARNISHES. CHINESE LACQUER. 259 

"Euphorbium." None of the Indian varnish-trees west of the Ganges 
yield as white or pure a lacquer as those in CJhina or Japan. 

A species of varnish-tree that grows in India was thought to be the 
veritable Anacarde, but it is entirely different from the Japanese 
''urusi^' variety. 

No attempt to cultivate any of the varnish-trees on a commercial 
scale has been made in either Europe or America. The manufacture 
of lacquer and lacquer-ware is one of the most important industries 
of China and Japan. It seems natural that if the largest users of 
varnish in the world depend almost solely upon these natural products, 
their cultivation in America is well worth trying. 

In China the Rhics vemicifera, or varnish-tree, is called Ch'i-shu 
{T^-chou), also Li-tschi, It grows wild in the province of Fingo 
and on the island of Tricom, and is cultivated in the mountains of 
Hupeh and Seechwan, but the best varieties are found in the province 
of Jamatto, where it is cultivated extensively. It is probable that 
the Mingpo and Foochow varnishes, as well as the Hupeh, are from 
the Rhus vemicifera A varnish-tree growing in South China differs 
from the above variety, but is not well known at present. 

In China the Rhits vemicifera grows about 15 to 20 feet in height, 
seldom reaching one foot in diameter, and has but few branches. 
The bark is white, knotty, and peels readily. The wood is fragile, 
resembling the willow; the pith is very abundant. Its leaves have a 
mild taste, and when rubbed on paper, dye it a dull black. The flowers 
are greenish yellow, and have an odor resembling orange-blossoms. 
The fruit is of the size and shape of chick-pea, and at its maturity is 
very hard and of a dirty color. The seed furnishes an oil and wax 
which are extensively used. 

From the berries of the Rhus vemicifera, Rhus succedabay and 
other related Chinese and Japanese species, a vegetable tallow is 
extracted and used for candles. The wax is exported in large quan- 
tities to Great Britain and the United States for an adulterant or 
substitute for beeswax. 

The general composition of crude lac is lac acid (a resinous acid, 
soluble in ether), 60 to 80 per cent, a gum 3 to 6 per cent, a nitrog- 
eneous substance resembling albumin 1.7 to 3.5 per cent of a volatile 
acid, and water, which are driven off in the preparation of refined 
lacquer. The color of the lacquer is a light yellow or brown, according 
to the season in which the tree is tapped. 

The Chinese crude lac is collected and purified in the same way 



260 NATURAL VARNISHES. CHINESE LACQUER. 

as in Japan. In both of the processes for its collection and refining 
great care is necessary. The poisonous element in the lac, whether 
inhaled or in contact with the flesh, produces what are known as 
varnish boils, accompanied by an intolerable itching and burning 
sensation, similar to that produced by the poison-ivy. They are 
difficult to heal, and resemble the effect of nitric acid on the flesh. 

The Chinese and Japanese use lacquer as a varnish or vehicle 
for colors on all kinds of household utensils, also for the inside and 
outside coatings on their buildings. Lacquer as a vehicle can 
be used for all colors except a pure white and some of the lighter 
shades of other colors. It is applied to wood, porcelain, and metals, 
and forms a hard resinous surface, highly lustrous, practically insolu- 
ble in boiling water, alcoholic liquids, alkaline and acid solutions, unless 
in a highly concentrated form. Applications of lacquer to the under- 
water surfaces of a number of Japanese war vessels for both anti-cor- 
rosive and anti-fouling coatings have been very successful. The 
coatings, after a sea-duty of four years of the vessels to which they 
were applied, showed no signs of either fouling or corrosion. 
Applications of other anti-fouling paints of all characters over 
lacquer coatings were failures, the urushic add of the lacquer attack- 
ing the metallic base of the foreign anti-fouling paint, resulting 
practically in the destruction of both. 

The best results for under-water marine work with lacquer is had 
when the first coating is a heavy one and almost pure lacquer. The 
succeeding coats can be thinner in body and contain either a pigment 
or some inert substance to give body. Mica, graphite, lampblack, 
etc., have been used experimentally with success for these secondary- 
coatings. 



CHAPTER XXVII. 

DECAY OP PAINT. 

Rust proceeds solely from the action of an acidulated moisture 
upon a bright or clean iron siuiace, and is probably only a point 
at its inaugural. The affinity of the iron for the oxygen in the acidu- 
lated moistiu*e of the air or water in the oil, or from other sources, 
is greater than its bond with the hydrogen as water (H,0), the decom- 
position ensuing releases the hydrogen, which is sixteen times the 
volume of oxygen united with the iron to form hydrated ¥efi^ or red 
rust. The hydrogen, from its light specific gravity, in its effort to 
escape into the air pushes up the overlying paint coating, increases 
the area of the affected part, cracks the coating in its exit, moisture 
enters again, and corrosion is master of that location. The rust 
which has thus been formed is hygroscopic and carries 24 per cent 
of moisture as it forms. This moisture never dries out under any 
atmospheric heat conditions, but is ever ready for a chemical decom- 
position; the hydrated red rust, being nearly two times the volume 
of the iron from which it is formed, adds its efforts to the free hydrogen 
to push up the coating and form a blister and crack in the coating. 
How energetic this mechanical action due to corrosion is, may be 
observed on the ordinary cast-iron hand railings for fences and out- 
side steps of New York City and other city houses, which in hundreds 
of instances are split for more or less of their length. Cast-iron 
water or gas-pipes, with bell and spigot joints, are frequently made 
with rust joints. They almost invariably burst the bells by the swell- 
ing of the iron cement used to make the joint. 

The cut. Fig. 35, shows a section of a well-known railway viaduct, 
the iron construction having been painted over mill-scale, or in the 
condition the material left the rolling-mill and workshop. It had 
received the usual treatment given by contracting engineers to remove 
the mill-scale preliminary to painting. 

Many sections of the New York City and other elevated rail- 
ways, also the Brookl3m Suspension Bridge trusses show mill-scale 

261 



262 DECAY OF PAINT. MILL-SCALE CORROSION. 

corrosion to an equal extent. Fig. 36 shows the mill-scale c 

on one of hundreds of New York elevated railway columns, orig- 



FiG. 36. — Mill-scale corrosion, Phfpnix column. 
inally painted with red lead. The corrosion now in progress is 
strong enough to break through and cast off six or more paint coat- 



DECAY OF PAINT. MILL-SCALE CORROSION. 263 

ings that have been applied over the red lead since the columns 
were placed in position. 

The corrosion existing in Bninell's* tubular iron bridge over the 
St. Lawrence River at Montreal, Canada, had proceeded to so great 
an extent as to require the removal of the whole structure, it being 
impossible to repair it. The efforts to replace the cross floor-beams 
supporting the rail stringers resulted in loosening every rivet in the 
neighborhood of the repairs. Pitting around the heads of the rivets 
had proceeded so far and deep that it was impossible to cut them 
out without loosening every contiguous rivet. 

This bridge had been kept well painted with iron-oxide and some 
experimental paints applied coat aft«r coat. These coatings when- 
ever removed, or that fell off diu-ing the periodical clean-up, or attempts 
to repair and paint the structure, showed the several coatings of mill- 
scale, paint, and new rust formations as plainly arranged as the leaves 
of a book. 

Fig. 37 shows a similar state of corrosion. 



Fio. 37. — Corrosion of steel girder. Woshinfrton Street railway bridge, Boston, 

In 1879, Sir Nathan Barnaby statal as the result of hi.s obser- 
vations of ships' metal in the English naval stations, that when the 
mill-scale was left upon the plates, angles, and other parts of the 
ship, its effect upon the neighboring bared metal was as strong and 
(â– ontinuou.s as copper would be. 

• Transactions .American Society Mechanical Engineers, Vol. XV, ]S94, 
paper number 628. p. 410. 



264 DECAY OF PAINT. MILL-SCALE CORROSION. 

In 1887, Mr. Rialton Dixon gave before the Institute of Naval 
Architects his experience as to a vessel built entirely of steel some 
eight years before, which was found to be greatly corroded in the 
bunkers and water-ballast chambers near the engine room and boilers. 
Some of the angle-irons had entirely disappeared, and the tie-plates 
were eaten away in holes. This action could be traced directly to the 
presence of mill-scale, and whether the surfaces were coated with 
paint or cement or not, the corrosion was always present upon those 
plates and angles that had mill-scale upon them, and tm« absent in 
those free from it. The presence of the paint or other coating retarded 
corrosion only in a minor degree by preventmg moisture from reach- 
ing the metal covered by the mill-scale. 

In 1882 Mr. Farquharson, on behalf of the English Board of 
Admiralty, conducted a number of very exhaustive experiments at 
the different naval stations to test the action of mill-scale on ships' 
metal. The result was to establish beyond dispute that, first, 
no pitting occurred on mild steel when freed from miU-scale; second, 
that the loss in weight from corrosion of clean mild steel and clean 
iron did not differ much ; third, that the action of mill-scale is con- 
siderable and continvxms, and equal to a similar amount of copper in 
its corrosive action on metal covered by it. Since these experiments 
the Admiralty have never wavered in their practice of having all 
of the ships' metal pickled to remove the mill-scale, whether it is 
to be covered by paint or cement, or to be galvanized. 

Destructive Agents of Paints. 

Pure water is a greater destructive element to an oil coating 
than solutions of sal ammoniac, chloride of magnesiimi, common 
salt, or natural sea-water, if free from sewage, all of which are agents 
of destruction. The decay of a paint is hastened by mechanical action 
if the water, either fresh or salt, or the other solutions, are in motion. 
Ordinary commercial oil coatings are destroyed by diluted muriatic 
and nitric acids, alkaline liquors, ammonia, sulphide of ammonium, 
soda, caustic alkalies, and alkaline solutions of coal ashes, clinkers, 
cinders, soot, etc. Diluted sulphuric acid does not materially affect 
an oil coating. All gaseous acids destroy the coating quicker than 
the acids in diluted aqueous solution, the destruction being in all 
cases hastened by heat or motion. Hence, to determine the probable 
protective value of any paint or other coating, it is necessary to know 
the detrimental influences to which it is to be subjected. 



DECAY OF PAINT. 265 

Changes in Paint Coatings. 

A coating of paint appears to be a very simple thing, as it is, when 
applied to a house or barn and both are left to their fate, but when 
applied to an important engineering structure, with all the vicissi- 
tudes of service in the extremes of heat and cold, sunshine and storm, 
atmospheric and other gases from natural or manufacturing som-ces, 
from corrosive liquids and solids, it is a different matter, and requires 
more engineering experience to select, more chemical knowledge to 
compound, and more technical details to get the right thing in the right 
place at the right time, in the right manner, and in the right amount 
than the general run of master painters do or can give to the subject. 
If the influences to which a coating of paint is to be subjected are 
known, it can generally be determined in advance whether it will 
be durable. For instance, zinc white or oxide (ZnO, specific gravity, 
5.42) applied as an external coating absorbs carbonic acid from the 
air and some moisture, changing to a carbonate of zinc (ZnCO,, 
specific gravity, 4.44). During this change there is an increase in 
volume from 14.9, as an oxide, to 28.1 as a carbonate. This change 
from an oxide to a carbonate is a chemical one, and occurs during 
the process of dr>ung, but the change in the volume of the two substances 
exerts a mechanical action also in the atoms of the pigment, not only 
to disrupt them and leave them loose and easily carried away by 
the wind, rain, etc., but cracks and loosens the oil vehicle in which the 
pigment is embedded as well as its bond to whatever surface it covers 
But if the zinc-oxide coating is applied in a closed room, though the 
air contains the same amount of carbonic acid, or even more than 
the external air, the oxide does not change to a carbonate, as the 
necessary moisture is lacking; hence zinc oxide for internal coatings 
is durable, but for outside coatings is perishable. 

Red lead (PbjO^, specific gravity, 9.07) remains unchanged under 
ordinary atmospheric conditions, but if the air contains hydric sul- 
phide, as it does in many manufacturing establishments and towns, 
to a notable extent, it will by an inexorable chemical law change 
the oxide to a sulphide of lead (PbS, specific gravity, 7.13), and 
this chemical change (usually denoted by the blackening or discolora- 
tion of the coat) will also be accompanied by an increase in volume 
of the sulphide of about 22 per cent, this increase acting mechanically 
to disturb the bond between the pigment vehicle and surface coated. 

The addition of carbonate of lime (chalk) to an iron-oxide pig- 



266 DECAY OF PAINT. 

ment, whether made from the iron ore, or from calcined copperas 
(FeSO^THjO), to neutralize the sulphuric acid developed in the cal- 
cination of the copperas or roasting of the ore (as heretofore noted), 
is another instance in which an inexorable chemical change in one 
of the pigment's loose substances is accompanied by a change in its 
specific gravity, its corresponding change in volume, and a mechan- 
ical action to reinforce the chemical action due to the raw-oil vehicle 
loaded with its charge of driers, whose function is io either decom.- 
pose or consume by a slow combustion the ^^ mucosities" in the oil 
while attempting to dry. All these instances are similar in effect 
to what would occur in the plastered wall of a building if the mortar 
used in it, when partially dry, should begin to increase in volimie 
to the amounts as given above. Other instances could be cited, but 
these show that the pigments of the coating can be so chosen as to 
preclude the destruction by them of the coating, but that it is almost 
impossible to guard the vehicle from the injurious influences inherent 
in the composition of the pigment, that is changed in character, after 
its application, by chemical laws. Hence the absokite necessity 
that an order for a protective paint should include the conditions 
it is to be subjected to. 

In addition to the preceding remarks upon iron oxide, graphite, 
and other paints, and the several tests given in detail of a few of 
the many paint compounds, it may be noted, viz. : 

All pigments * can be grouped into three classes, according to their 
affinity for linseed-oil. 

First. Those that form chemical combinations called soaps and 
are generally the most durable. They consist of lead, zinc, and 
iron bases, of which red lead combines with the oil to the greatest 
extent ; next, the pure carbonate white lead made by the "Old Dutch 
Process," followed by zinc oxide and iron oxide, Turkey umber, yellow 
ochre; also, faintly, the chromates of lead, chrome-green, and chrome- 
yellow. 

Second. Pigments of this class, being neutral, have no chemical 
affinity for the oil; they need large amounts of driers, either com- 
bined with and carried by the oil, or as free driers. They include 
all blacks, graphites, slates, slags, vermilions, Prussian, Paris, and 
Chinese blues, terra de sienna, Vandyke brown, Paris green, verdigris, 
ultramarine, carmine, and madder lakes. The last seven are trans- 

* " Pigments." English Encyclopedia of Painting, 1880. 



CATALYSIS IN THE DECAY OF PAINT, 267 

parent colors, and are better adapted for varnish mixtures and glazing. 

Third. Pigments of this class act destructively to linseed-oil. 
They have an acid base (mostly tin salt, hydrochloride of tin, and 
redwood dye) which forms, with the albuminous and gelatinous 
matters in the oil, a jelly-like compound that does not work well under 
the brush nor harden sufficiently, and can be used in a varnish for 
glazing only. Among the most troublesome are the lower grades 
of so-called carmines, madder lakes, rose-pinks, etc., which contain 
more or less acidulous dyes, forming with linseed-oil a soft paint, 
that dries on the surface only and can be peeled ofif like the skin of 
ripe fruit. 

"Catalysis " is a term introduced by Berzelius, and by him applied 
to the changes that sugar solutions undergo in the process of fer- 
mentation, and now used to denote the changes that certain substances, 
by their mere presence, effect in other bodies without themselves 
undergoing any apparent change. Catalytic action is a potential 
agent in the decay of paint coatings, and manifestly has not received 
the attention from paint chemists and compounders that its marked 
action on the life of a coating warrants. The present efficiency of 
the incandescent gaslight is wholly due to catalytic action between 
the substances that compose the mantle when excited by the com- 
bustion of the gas. In the development of this light all of the rare 
mineral oxides and metals and the oxides of the baser metals, chro- 
mium, alumina, cobalt, manganese, nickel, and iron, when associated 
with thoria in the mantle, have been found to act as catalytic agents 
to carry, condense, or absorb oxygen, that increases the flame tem- 
perature of the mantle and consequently increases the light. This 
flame temperature in some cases reaches the point where volatiliza- 
tion of some of the baser metals and oxides ensues. Charcoal, pow- 
dered glass, porcelain, flour spar, crj'^stallized quartz, pumice-stone, 
and other kindred substances are also found to act as catalytic agents 
in combustion, but do not develop so high a flame temperature in 
the mantle as the other substances above noted. 

Combustion of any substance may be quick and attended by a 
high temperature, as in the case of the incandescent gaslight, or it 
may be of low temperature and extend over years of time, but the 
amount of heat evolved from the destruction of the substance and 
the resultant products of combustion, or decomposition, are the 
same in all cases, even if the physical effects are apparently different. 

Nearly every substance in a paint coating has been found to be 



268 CATALYSIS IN THE DECAY OF PAINT. 

catalytic to some other substance, either in its own class as a so-called 
inert mineral pigment, or in the chemical class of oxides having a lead, 
zinc, iron, or other metalUc base. Individually, they may be appar- 
ently unaffected by long exposure to the air while in their loose state 
or in packages or bulk ; but when mixed together, they take up moist- 
ure or oxygen to a greater or less degree, either by absorption in 
mass or by condensation upon their surfaces, and catalytic action 
ensues. The oil vehicle and driers are catalytic of themselves, and 
when mixed with the pigments act more energetically as carriers 
of oxygen even when the coating is apparently dry. In all pigments 
and vehicles, the one that is the most refractory, or that is the most 
resistant to oxidation in whatever form the oxygen may be presented, 
is the one that acts the part of the thoria in the gaslight mantle, 
becoming the negative or non-consumable substance, that, though 
excited to a greater activity by the presence of the other substances 
in the paint compound, retains its resistance to a change the longest 
at the expense of the other associated substances. Thus far, lamp- 
black and graphite, in their subdivided form as pigments, appear 
to be the only substances not subject to catalytic action, or if it is 
present it is so weak that the life of the coating is not materially 
aflFected from this cause.* 

Caustic Action of Mortar upon Paint. 

An examination (1901) of iron floor-beams taken out after an. 
exposure of about forty years showed that the beams originally 
were particularly well painted and laid in a location where only 
the dry warm atmosphere of a residence reached them. The paint 
coatings had been thoroughly destroyed by the caustic action of 
the lime mortar used to turn the brick arches in which the beams 
were embedded. Corrosion was well established in every inch of 
their surface. Had any moisture, as in the case of the Times build- 
ing, reached them, their condition would have been fully as bad. 

The iron beams supporting the sidewalks laid about forty years 
ago in New York City, that were removed for the Rapid Transit 
Tunnel work, invariably show deep corrosion from the destruction 
of the paint coatings from the caustic action of the lime mortar. 



* A fuU list of the series of electro-chemical elements having a metallic base 
and entering into the composition of pigments is given in Chapter XXXVI. 



CAUSTIC ACTION OF MORTAR UPON PAINT. 269 

also, that the dried mortar is not a protection from corrosion, but a 
promoter of it, if moisture or air can reach the surface so covered. 



Fig. 38. — Corrosion of eidewalk iron bcama. 

Hydraulic, also quicklime mortar, only prevent corrosion so far 
as they are free from mill-scale and continuously dry to exclude 
the air. The paint coating when burned by the caustic action of 
mortar or cement, adds no material period to the life of the iron 
and except for appearance and protection during conatruction, might 
be left off. (See Chapter XV.) 

The modern hollow tiles used for floor arches and building partitions 
with their advantages over brickwork, do not remove the cause of 
the corrosion of any iron that they may be in contact with. 

Gypsum, while not caustic, is hydrometic, and the continual pres- 
ence of moisture is fatal to ferric bodies; besides, it is not always 
free from caustic substances developed in the calcination of it. 

The following cement for the levelling, bedding, and in contact 
with metal work, is recommended. The cement hardens like stone, 
is impervious to water, and can be applied by a trowel from a mortar- 
board, over walls or to lay brick wherever mortar can be used. It 
is made from marble dust (from marble sawing or pulverizing mills) 
mixed, viz.: 

Pulverized marble 62 per cent. 

Sharp silicious clean sand 35 " " 

Litharge 3 " " 



270 CAUSTIC ACTION OF MORTAR UPON PAINT. 

These proportions can be varied somewhat without injury. Too 
much Umestone impairs the hardness; too much sand makes the 
cement porous. When the mastic is to be used, for every 100 parts 
of such mixture, 7 parts of Unseed-oil are required to bring it to a 
good trowel paste. The oil can be either raw or boiled, according 
to the time of drying required. The surfaces to which it is to be applied 
should be dry, clean, and preferably coated with linseed-oil or a good 
carbon paint, before the appUcation of the cement. 

A refined bitimien coating applied to the bright metal, hot, has 
proven to be the best of coatings for ironwork laid in cement, mortar, 
or concrete, to correct the caustic action of them. 

The metal work of the movable. dam at Lake Wennibioskish, 
Minn., constructed in 1899-1900, was cleaned bright by the sand- 
blast and then painted three coats of Edward Smith's Co.'s Durable 
Paint, applied one week apart, each coating- being thoroughly dry 
before the application of the next. Observation of the paint in 
1901 showed that the coatings had been completely killed and ab- 
sorbed wherever the painted metal was embedded in the concrete. 
The metal was as clean as before painting, with a slight discoloration 
of the surface of the concrete from the paint absorbed. The metal 
exposed, however, did not show the same tendency to rust quickly, 
as before the application of the paint, on the short exposure before 
again being put in place. The surfaces not in contact with concrete 
were in good condition* 



CHAPTER XXVIII. 

SAND-BLAST AND PICKLING PROCESSES. 

The sand-blast is the most satisfactory and simplest method of 
cleaning all surfaces for painting, whether at the shops or in situ. 



Fig. 39. — Sand-blast appsratua. 

The invention of the sand-blast is due to General Benj. Tilghman, 
and was patented October 18, 1870, No. 108,408, but has smce 
expired. There are some patents for sand-blast apparatus of subse- 
quent date, issued to other parties for improvements relative to 
portability, clogging of the sand in the case, etc., still in effect. 

Fig. 40 shows a portable sand-blast apparatus used by Mr. Geo. 
W. Lilly, C.E., for cleaning railway viaducts in the city of Colum- 
bus, Ohio. 

The principal features of the sand-blast consist in the use of 
compressed air at a pressure of from 15 to 25 pounds per square 
inch, discharged through one or more chilled-iron or hardened-ateel 
nozzles ,^ inch in diameter, which are directed upon the work to be 
cleaned. By suitable devices, into this current of air, dry sharp sand 



272 



SAND-BLAST PROCESS. 



or coarsely powdered quartz is fed at the rate of about 10 cubic feet 
per hour for each nozzle, which discharges about 120 cubic feet of 
free air per hour, or about 1 cubic foot of sand to 1000 cubic feet 
of free air per hour. The nozzles wear rapidly and require frequent 




Fio. 40. — Portable sand-blast machine. 

renewal, but they are of small or minor expense. The 2 or 2J-inch 
diameter armor-clad rubber hose that conducts the compressed air 
from the air-receiver to the place of work being soft and elastic, is 
comparatively little affected by the current of sand and air, unless 
the air is hot; hence methods to cool the air before it reaches the 
leading hose are necessary. Four such nozzles that gradually wear 
to |-inch diameter and then discharge 200 to 240 cubic feet of air 
per hour, require an air-compressor of 20 inches diameter for steam 
and 22-inch air-cylinders, by 24-inch stroke or approximate sizes, 
also a 150 H.P. boiler. 



SAND-BLAST PROCESS. 273 

The abrading material, when used, must be thoroughly dry, and 
can be used four or five times over, or until it is broken into a powder 
too fine, or becomes too dirty to be effective. 

For cleaning ships' bottoms in dry dock, where the rust and 
paint coatings are generally thick and somewhat softened by the 
water, about -^ square foot of metal is cleaned per minute, or 48 
to 60 square feet per hour. This costs about 3 cents per square 
foot of surface, as the waste of sand is greater and the work cannot 
be done so advantageously in a dry dock as in a shop. 

On the New York Elevated Railway Station at 156th Street an 
average of 80 square feet per hour was maintained for a number of 
months in removing a hard coating of old paint and rust to the bright 
kon. The loss of time in changing nozzles and shiftmg scaffolds 
was about one hour per day per nozzle. The labor account was 
one man to hold and direct each nozzle; one man to attend to two 
sand-boxes, and one man to clean up and supply sand for the four 
nozzles or seven men per corps the men shifting their scaffolds with- 
out other aid. The four-nozzle plant for bridge or viaduct clean- 
ing will clean about 2500 square feet of surface per eightrhour day 
at an expense of about $20 for all iten:s, except the man and coal for 
the compressor, or 8 cents per square foot, which would be modified 
by the amount of cleaning for each structure. 

Removal at the shop of mill-scale and dirt is done at the rate 
of 4^ to 5 square feet of smf ace per minute, or 270 to 300 square feet 
per hour per nozzle, or about i cent per square foot of surface. With 
an organized corps and plant, the cost of cleaning surfaces need not 
exceed J cent per square foot of surface, large or small, or about one- 
third the labor-cost of the painter on a first-class coat of paint, and 
requires about the same degree of skilled labor as painting. 

The metal-work of the movable dam at Lake Wennibioskish, 
Minn., erected during 1900, was cleaned bright by an extemporized 
sand-blast. A hoisting-engine run backward furnished the com- 
pressed air, an old steam-boiler was used for an air-receiver, gas- 
pipe for nozzles, and garden hose for leaders, etc. 

Mr.W. C. Weeks, C.E.,* reports "that four laborers and one engineer 
in charge of the apparatus cost for labor $9.22 and $2.50 for fuel. On 
general surfaces, 40 square feet per hour were cleaned, using two nozzles, 



* Engineering Record ^ May 4, 1901. 



274 SAND-BLAST PROCESS. 

or a cost of $0,036 per square foot. On plates and large straight sur- 
faces, 90 square feet per hour was the usual rate of cleaning, which 
cost $0,016 per square foot." 

In a number of United States navy-yards, with well-equipped, 
permanent, and fairly perfect sand-blast plants, the cost of cleaning 
averages i cent per square foot of surface. This for plates ^-inch 
thick is 98 cents; for J-mch plates, $1.95 to $2.00 per ton. To 
sand-blast 7-inch I-beams weighing 17.5 pounds per foot costs 
$1.35 per ton; 12-inch I-beams, weight 50 pounds per foot, cost 80 
cents per ton. 

The average cost for cleaning plate-girder bridges in situ is proba- 
bly $1.00 per ton of metal. For truss and lattice-iron bridges the 
cost of sand-blasting ranges from $1.00 to $1.75 per ton. 

The United States Army Engineer Corps cleaned 50,000 square 
feet of steel lock-gates and other metal on the Muscle Shoals Canal 
during 1898-99 from a temporary floating plant. The cost of all 
items, was 3 cents per square foot, and the new coat of paint cost 
2.88 cents per square foot. 

Mr. Geo. W. Lilly, C.E., reports the cleaning by sand-blast of a 
number of railway bridges and viaducts in the city of Columbus, 
Ohio. The work was done under exceptionally adverse circum- 
stances, but indicated that 8 cents per square foot covered all the 
expenses. For cleaning a viaduct over the Little Miami Railway, 
containing 25,000 square feet of surface in a confined location where 
the cleaning was interrupted by the train-service that frequently 
amounted to one-fifth of the working hours, the cost of the work, 
including flagman, sand and drying, compressed air, and all other 
expenses, including the labor, was 3.04 cents per square foot. On 
the best days, in a favorable location uninterrupted by the trains, 
1227 square feet were cleaned per day at a cost of 1.23 cents per 
square foot. On a plate girder containing 3727 square feet, the 
cost was 2.37 cents per square foot. 

The pressure of air ranged from 25 to 38 pounds, averaging 33 
pounds per square inch. The nozzles wore rapidly. They were 
of i-inch extra heavy wrought-iron gas-pipe, about 2 feet long and 
lasted from 3 to 5 hours each. 

The expense of handling structural metal at the shops after machin- 
ing preparatory to sand-blasting it, ranges from 40 cents to $1.50 
per ton according to the weight and character of the pieces and facili- 
ties of the plant. 



CLEANING PROCESS. 275 

A recognition of the fact that structural steel is a perishable 
material, requiring thorough protection from corrosion during all 
the stages of its manufacture and use, should be required of every 
engineer, and the subject should form an important part of his edu- 
cation. There is no part of structural engineering needing a more 
thorough reform in both spirit and practice than this one. 

The apparent indifference regarding the future fate of steel mate- 
rial, after it is in location, is probably due to the mistaken economy 
of the engineer corps and the proprietors, on account of the added 
cost of properly cleaning the metal. If cleaning is necassary, as engi- 
neers and all admit, it should be specified in the contract, properly done, 
inspected by a competent person, and paid for like any of the other 
processes, and the penalty for its non-fulfilment be as strictly enforced 
as for a badly driven rivet or poorly machined or fitting part. The 
above deficiencies are readily detected and can be corrected, but the 
poorly cleaned surface escapes notice and is readily put out of evi- 
dence by the handy paint-pot. 

It is the imperative duty of the engineer in charge of structural 
work to require his inspector to perform his duty at all times so that 
a radical change shall be had from the present practice of cleaning 
and painting ferric metal, the corrosion of which is now too much in 
evidence. 

At a late meeting of an engineering society, the protection of 
ferric structures from corrosion was under consideration by oral 
discussion and correspondence. The cleaning of the surface of steel 
to the absolutely clean metal by some method, preliminary to the 
immediate painting of it under cover, was unqualifiedly endorsed. 
And yet within the limits of a ten-mile circle from the engineers' 
meeting there were many thousands of tons of ferric structural mate- 
rial in process of erection by the engineers represented at the meeting, 
and scarcely a ton of this material had received any other cleaning 
than that from a putty knife or a whisk-broom. The quality of the 
applied paint in many cases was as deficient as the cleaning. So 
much for theory versus practice. 

Pickling the metal instead of sand-blasting it is more practised 
bv European than American engineers, especially for structural work. 
Mill-scale is readily removed by immersing the metal in a hot dilute 
solution of sulphuric acid. Generally, 6 to 12 minutes suffices, using 
a 25 to 28 per cent acid solution. A 10 or 12 per cent solution is 
effective, but requires more time in the bath. The stronger solutions 



276 PICKLING PROCESS. 

are recommended. They are equally as safe tx) handle and should 
be applied hot if possible; the latter quality is best for removing the 
scale. 

In foreign navy-yards, 9 to 10 per cent hot solutions are used, 
the metal remaining in the bath five or more hours, according to the 
quality of the scale. This requires a large pickling plant and has no 
other advantages. When appearance or test shows the scale is loos- 
ened, the metal is removed and well washed by a copious and strong 
jet of water under 75 poimds or more pressure. 

Soaking the metal in baths of still or light run n ing water does 
not thoroughly remove the acid. The still-water bath is the cause 
of the failure of tin-plate. 

Pickled metal is liable to become coated with a tough, gummy 
substance, quite difficult to remove, except by the friction from a 
strong jet of water. Arsenic in the sulphuric acid made from pyrites 
also adds to the gunmiy deposit precipitated on the metal. Acid 
free from arsenic should be specified for the pickle. The gunmiy 
deposit prevents the paint from bonding to the metal, rendering it 
liable to peel. 

After the metal has been washed by the jet of hot or cold water 
it should be immersed in a bath of hot lime-water and be left in it 
long enough to reach the temperature of the bath, in order to neu- 
tralize any of the acid not removed by the water jet. It is then 
removed and dried, preferably in an oven. The coating of lime left 
upon the metal can be easily brushed off, leaving the metal clean and 
bright, which will show evidences of rust in an hour if not painted 
immediately. 

Muriatic acid is sometimes used in place of sulphuric acid for the 
pickle. It is not as effective as sulphuric acid, it costs more, and the 
gummy coating formed by the pickle is more difficult to remove, 
requiring a hot alkaline or caustic-soda bath, instead of lime, to remove 
it. A solution of sulphate of zinc is effective for the removal of this 
gummy coating. 

Obviously a pickling plant requires a larger space for an equal 
cleaning capacity than a sand-blast, and is not so convenient to use 
at all seasons of the year, and both are impractical to use in the con- 
struction tool shops. 

The labor and material accounts for a pickling bath for all sizes 
and weights of ordinary structural steel is about 25 cents per ton. 
The labor account for moving the pieces into and out of the pickle, 



GASOLINE PROCESS FOR CLEANING METAL, 277 

cleaning baths, and ovens, will be from 60 cents to $2.00 per ton, 
or rather more than is required for a sand-blast, as the several pieces, 
though of the same weight and character, have to be moved more 
frequently. 

Steels high in carbon, or cast-iron articles, are diflScult to pickle, 
as a film of graphitic carbon forms on the surface of the metal, which 
mixes with the gummy deposit from the acid bath, and requires 
considerable labor and care to remove. 

When the sand-blast or pickling process is not available, mill- 
scale, rust, and old paint coatings are removed from works in situ, 
by the gasoline burning torch, followed closely by the scraper and wire 
brush. 

The cost of the burning process is so closely connected with the 
painter's labor as to be diflScult of separation, but a quart^buming 
torch will bum 3^ hours, and one man can saturate rust-spots and 
burn off from 80 to 100 square feet of surface per hour, at a cost of 
iV to yV cent per square foot of surface, leaving it ready for the painter. 

A modern parlor, or sleeping-car, 65 to 70 feet long, requires three 
gallons of gasoline to burn off the outside paint coating, and about 
four days of time, for one man to use the torch, followed by two men 
two days each, to sandpaper ready for the painter, or a total cost of 
45 cents for the gasoline, and $15.00 for the labor, or -^ to 1 cent 
per square foot of surface. 

Care is required in the use of gasoline, either for the torch, or for 
saturating the old paint, as explosions and serious burning of the 
workmen, and fires, are frequent. Insurance companies forbid the 
use of either the torch or fluid for the removal of paint or rust in any 
building covered by their policies. 

Any material that can be inclosed in a chamber or iron casing 
and subjected to the action of a bath of low-pressure steam for 20 
to 25 minutes will have the old coatings softened, when they can be 
easily scraped off. This is to be followed by a thorough washing 
with soap and water, and rinsing. A pair of locomotive driving- 
wheels required 30 minutes to scrape and wash after steaming. The 
total cost being about one-third to one-half that required for the 
usual caustic-soda application. 

Many railway repair shops use the following mixture for the 
removal of old paint. It has no action upon rust. Caustic soda 
and sal-soda, each 30 pounds; mix with 3 pounds of strong ammonia 
diluted with 30 gallons of water. To the above, add a mixture of 



278 OTHER PROCESSES FOR CLEANING METAL, 

30 pounds of finely ground quicklime in 5 gallons of water, and 3 to 4 
pounds of melted laundry or soft soap. The two mixtiires added 
together, when cold, should be of the consistency of putty. It is 
applied by a trowel or stiff 4-inch flat brush in successive coats about 
J to i inch thick. Care must be used in mixing the lime. A stirring- 
paddle should be left in the tub to form a vent to prevent the caustic 
mixture from blowing out. 

The cost to remove the paint from a pair of locomotive driving- 
wheels by this mixture is 65 cents for material and 15 hours of labor 
at $2.25, or a total of $2.90. Careful washing with hot wat^r to remove 
all traces of the caustic-soda paste is required, as for all strong alka- 
line mixtures. 

Wooden surfaces treated with caustic-soda compounds to remove 
paint or varnish are injured by the raising of the grain of the wood, 
which cannot be restored by sand-papering. The parts so treated 
show spotted; even a staining-coat will not cover them unifonnly. 
Fine woods are injured the worst. 



CHAPTER XXIX. 

FERRIO-PAINT TESTS. 

Objection is made by some engineers and paint manufacturers 
to the inunersion methods of testing paints; that they do not meet 
the actual conditions of coatings exposed to weather; that a ferric 
structure is not always wet, but wet and dry, with more dry hours 
than wet, etc. This would depend altogether upon the location of 
the structure; in many instances there might be more wet or damp 
hours than dry ones. A fog or long-continued sweat is more destruc- 
tive to a paint coating than a passing storm. But the plain fact 
remains that these tests are all competitive as between different 
commercial paints, and imder xmiform conditions. The trial given 
one paint is given to all; the few successful ones are the better ones 
to select from to base any subsequent improvements or experiments 
upon, or for use. The water-test settles the merit of a protective coat- 
ing in short order, and so soon as generally adopted by those ordering 
paints for the protection of ferric structures exposed to weather, so 
soon will the great majority of these patent paint compounds cease 
to vex the engineer with high claims and low performance. 

The nearer any protective coating approximates an enamel or var- 
nish, generally the more durable it will be. The Japanese and Chinese 
lacquers are varnishes, and dry better by the application of water 
than in dry air alone, and all compounded varnishes are hardened in 
the last stages of their drying by water. Lacquers when thoroughly 
dry remain unchanged for scores of years, when exposed to fresh or 
salt water either hot or cold or alternately wet and dry, or immersed 
for years. The coming ferric protective coating will probably be a true 
varnish with a carbon or graphite pigment. But it will be well to 
bear in mind that it will not be imperishable in exposed locations, 
and that its application and the preparation of the structure to receive 
it will require more attention than at the present time these matters 
receive; neither will it be a low-cost article. 

279 



280 FERRIC-PAINT TESTS. SMITH'S TESTS. 

Tests of Twenty-seven English Commercial Ferric Paints. 

In a paper * read before the Newcastle, England, section of the 
Society of Chemical Industry, Mr. Henry Smith, F.I.C., described 
a series of experiments upon the protective powers of twenty-seven 
different English commercial paints, as applied to ironwork in fifty 
separate instances. The methods of test were those devised and 
employed by Mr. Max Toltz, C.E., in a series of experiments upon a 
number of American commercial protective coatings for iron in 1897.t 
Three sets of bright and clean iron plates, all of the same size, were 
respectively coated with the several paints, in all cases furnished as a 
stiff paste, and when applied, were brought to the consistency of a 
paint by mixing with genuine boiled linseed-oil, capable of drying 
in seven hours under ordinary conditions of temperature, no driers 
being used. The first coat was allowed to dry thoroughly firm before 
the second coating was applied. When this also was firm and hard, 
one set of the plates was exposed to the weather, as in ordinary cases 
of painted structures. The other two sets were treated as follows: 
One set was simply to corroborate the results obtained from the 
other set, the results being practically identical in each case. Each 
painted strip was placed in a clean, wide-mouthed glass bottle, half 
filled with clean pure water. The bottles were not closed, but were 
protected from the entrance of dust and impurities while allowing the 
air free access to the painted plates. Several of the plates had com- 
menced to corrode in about a week. This was indicated by a cloudi- 
ness in the water, which afterward became further oxidized, and 
formed a red precipitate of ferric oxide, which subsided partly to the 
bottom of the vessel. After three months' exposure the plates were 
removed, and the liquid in each bottle, together with the sediment, 
was tested for the percentage of iron present in the form of rust. 

The figure given as denoting the amount of corrosion is less than 
the actual amount, as it does not include the portion that adhered 
to the plate, and was not scraped or brushed off, and would not drain 
off. In each case the weight of rust was calculated to pounds of rust 
per 1500 square yards of painted surface; the other figures give the 
percentage composition of the several paints by weight. 

* Engineer (London) and the American Gas Light Journal (New York), 
f September 4-20, 1899. Journal of the AssociaHon of Engineering Societies^ 
St. Paul, Minn., 1897. 



FERRIC-PAINT TESTS SMITH'S TESTS. 



281 



CONDITION AFTBR THREE MONTHS' EXPOSURE. 

Pounds of Rust from 1500 Square Yards of Surface. 

Corrosion. 

6 samples of red lead alone, or mixed with barytes 50% ; raw oil, 10.00% None. 



Red lead. . . 22.00% 
" " . . .33.33% 
3 samples zinc oxide 
Zinc oxide. .27.27% 
White lead, pure . . 
White lead. .53.78; 
" " 50.52%; 



barytes, 66.00% ; total, 88.00% ; 



(( 



n 



n 



it 
it 



58.80%; 

45.00%; 

63.63%; 

92.56%; 

barytes 40.33%; 

42.10%; 



it 



..92.13%; 



tt 



It 



tt 



tt 



tt 



Iron oxide, pale (50% Fe^Oa). 83.60%; 



12.00% 
7.87% 

10.00% Trace. 

..90.90%; boUed oil, 9.10% " 

" " 7.44% 75 lbs. 

..94.11%; " " 5.89% 80 

..92.62%; " " 7.38% 95 

16.40% 81 " 



it 



tt 



tt 



tt 



tt 



tt 



tt 



It 



It 



tt 



tt 



deep (96% Fe,OJ. .86.89%; raw oil, 13.11% 123 



tt 



Venetian red 7.55% ) 

Barytes 80.57% ) 

Iron oxide, medium color (94% 

Fefi^ 86.89%;. 

Iron oxide, extra bright color 

(90% Fe,0,) 82.35%; . 

Iron oxide, pure (90% FejO^) . 76.30% ; . 



tt 



..88.12%; 



tt 

It 

tt 
tt 



tt 

tt 

tt 
tt 



11.88% 123 " 



tt 



medium 12.30% ) 

Barytes 76.22% ) 

Indian red (70% Fe^OJ 82.35%; , 

Turkey red (95% Fe,0,) 81.16%; 

Iron oxide 27.03% ) 

Bar3rtes and calcitmi carbonate.62.52% ) 
Barytes (natural barium sul- 
phate) 88.00%; , 

Iron oxide, Venetian red 8.47% ) 



It 



.88.52%; 



tt It 



It 



• . civ. Do /q ; 



It 
It 

tt 



tt 
tt 

It 



tt tt 



It 



..87.27%; 



It 



It 



13.11% 134 

17.65% 137 
23.70% 160 

11.48% 244 

17.65% 227 
18.84% 262 

10.55% 398 



12.00% 156 
12.73% 118 



i 



II 



It 



II 



..86.07%; 



..92.39%; 



II 



It 



It 



It 



12.14% 242 



7.61% 266 



Barytes and calcium carbonate.78.80% 

Iron oxide 13.93% 

Barytes and calcium carbonate.60.00% , 
Rose pink (principally barytes) .12.14% ) 
Bar3rtes and calcium carbonate.80.56% 1 

Celestial blue 11.83% J 

Barytes and calcium carbonate.68.99% \ 

Ivory and carbon black 8.42% V 

Manganese dioxide 2.46% J 

Barytes and calcium carbonate.79.30% \ 

Carbon and bone black 4.35% j- 

Manganese dioxide 1.30% ) 

Drop black (charcoal black) . . .60.00%; boiled oil, 40.00% 260 

Flake graphite, pure 69.66%; raw oU, 30.44% 215 

Boiled linseed oil, pure 600 

Raw turkey umber 51.85%; raw oil, 48.16% 610 



ti 

tt 
tt 

II 

u 
u 

u 

n 

II 

tt 
tt 



. . 79.87% ; boiled oU, 20.13% 352 " 



..84.95%; raw oil, 15.05% 392 " 



II 
II 
It 
It 



282 FERRIC-PAINT TESTS. SMITHES TESTS. 

Twenty mixtures of barytes alone, or with calcium carbonate 
mixed with celestial blue, Prussian blue, chrome-yellow, raw sienna, 
Vandyke brown, Italian ochre, Brunswick and other greens, chromate 
of lead, English umber, Turkey umber, ultramarine, Chinese blue, 
burnt sienna, mixed with raw oil in proportions from 11 per cent to 
51 per cent of the weight of the paint; the corrosion in the order 
named above ran from 168 pounds to 441 pounds per 1500 square 
yards of surface. 

Except in the case of the blues, umbers, siennas, etc., where the 
pigment had but Uttle influence on the oil to resist decay beyond 
that inherent in the oil alone, the more separate substances that 
entered into the composition of the pigment, the more unreliable it 
became. A single exception is noted in the case of a Venetian red 
paint, made from barytes, calcium carbonate, and a small amount of 
iron oxide, that gave a better result than barytes alone, or when 
barytes was mixed with the other color pigments of much less specific 
gravity. Several substances in a composite paint are generally fatal 
to its protective qualities, no matter to what it is applied. The 
several atoms of these substances, even if uniformly distributed in the 
pigment in the process of grinding, bolting, and mixing (but they are 
not), will not retain their juxtaposition when mixed with the oil. 
The heavy atoms will sink, and there will be a marked difference in 
the coating spread from the top of the paint in the pot from that in 
the middle or bottom; the lightest and most perishable substances 
will get on the surface firs 

Barytes worked well with red lead and zinc oxide, there being but 
a small difference in their specific gravities as compared with barytes 
and the other color or base pigments. With white lead, as the per- 
centage of bar3rtes was increased, so was the corrosion. Aside from 
the reduction in cost of these lead and zinc pigments by the addition of 
barytes, there is no reason for its use, as the barytes alone did not 
give a satisfactory test. No doubt from the splintery character of 
its atoms, as has been before commented upon, it is wholly destitute 
of covering or coloring power. The vagaries of the iron-oxide paints 
in the varying proportions of the pigment and oil are noticeable, but 
not so marked as where barytes, one of the heaviest of all pigments, 
and calcium carbonate, one of the lightest, both classed as inert 
pigments, were mixed with the oxide, and fully sustained the pre- 
vious remarks upon the non-protective character of composite and 
iron-oxide paints. Boiled oil, in the single instance reported, proved 



FERRIC-PAINT TESTS. SMITH'S DISH-TESTS. 283 

superior to raw oil as a vehicle for the several iron-oxide paints in the 
ratio of one to nearly five. 

Smithes Di8h4e9t8 of Painta. 

A second series of experiments were made by the same experi- 
menter, and following the method of Mr. Max Toltz, C.E., to wit: 
A number of iron dishes five inches in diameter and one-half inch deep 
were scoured bright, and then coated with two coats of the several 
paints used upon the above-detailed iron plates and under the same 
conditions as to the composition and drying of the paints. These 
shallow dishes were filled with water and allowed to completely 
evaporate in the open air of the laboratory. This operation was 
repeated six times in the course of six months. Thus tested, the only 
paints which remained practically imaffected were red-lead and 
orange-lead paints, some of which, however, such as the "vermilion- 
ette" and scarlet-red paints, contained a' so a proportion of aniline 
colors, while two of the red-lead paints contained in the one case 
45 per cent of barytes and in the other 66 per cent. All the other 
dishes were more or less rusted, the order of merit of the better paints 
being as follows: 

1st. Zinc oxide. 

2d. Equal parts zinc white and barytes. 

3d. Zinc white, 3 parts; barytes, 7 parts. 

4th. Lithopone (a mixture of zinc sulphate^ zinc oxide, and barytes). 

5th. Pure white lead. 

6th. White lead, 5.37 parts; barytes, 4.03 parts. 

7th. White lead, 5.05 parts; barytes, 4.21 parts. 

All the other paints, thirty-six in niunber, proved inefficient. 
The first to show rust was that one painted simply with linseed-oil. 
The above classification of merit is by Mr. Smith, and, taken together 
with the detailed report of the glass-bottle test (before given), may 
be considered a fair representation of the protecti^ve qualities of the 
hundreds of commercial ferric paints foisted upon the market under 
various trade-mark names in the United States as well as in England, 
where the above experiments were conducted. 

Both the immersion- and dish-tests are very important for deter- 
mining in a relatively short time the weather-resisting power of a 
paint. If the coating is unable to resist the action of water or moisture 
in the form of steam, fog, or vapor from a tunnel or other confined 
space, it cannot be desirable for the protection of a ferric structure, 



284 FERRICJ'AINT TESTS. TOLTZ'S DISH-TESTS. 

or even a wooden one. The dish-test, probably, is the nearest to the 
actual condition which a paint must withstand. When the water 
in the dish is nearly evaporated, there remains in the circular seam 
of the bottom a film of water which contains the carbonic acid and 
the decomposing gases and dirt from the atmosphere, which act 
upon the paint in such a way that the coating at that part is soon 
permeated and rust forms. This action is more and more developed 
after each evaporation, and practically covers the whole dish in a 
short time. In actual service the same thing will happen. The 
corner of the dish finds its counterpart in every corner of a ferric 
structure where two plates, angles, or other parts join. Rust will 
commence at those seams and extend imder the paint, but will not 
show as plainly on a bridge-truss as on the small dish. 

ToUz's Tests of American Commercial Ferric Paints. 

The shallow-dish tests by Mr. Max Toltz, C.E. (before referred to), 
were made prior and during 1897, and extended over a period of 
from six months to two years. Without entering into as great detail 
as that quoted from Professor Smith, the deductions from his tests 
are in brief. Twenty-two different paints were submitted to test 
under the following classification: 

No. 1. True asphaltic varnish paints compotmded by heat in the 
same manner as a black baked japan, and practically of the same 
nature and comparable therewith. No corrosion reported after the 
dishes had been filled and evaporated naturally fourteen times. 

No. 2. So-called asphaltic varnishes, or paints of inferior qualities 
to the above No. 1, made from asphaltum dissolved in benzine or 
other volatile vehicle, but were not a true varnish. They contained 
about 43.5 per cent of vehicle and 56.5 per cent reported to be as- 
phaltum. As a rule they showed well in the beginning, but after the 
volatiles had evaporated, especially when subjected to a moderate 
heat-test, the coatings became quite brittle, were easily removed 
by abrasion, and did not protect the surface covered with them. 
Their composition varied in the several specimens tested. One sample 
analyzed had no asphaltum in it. Under test the dishes painted one 
coat showed considerable rust all over after the fifth exposure. Those 
painted two coats after the seventh exposure showed not much better. 
Generally, their reliability as protective coverings for ferric struc- 
tures is the least satisfactory of all paints. 

No. 3. Black-carbon paints, in which the vehicle was practically 



FERRIC-PAINT TESTS. TOLTZ'S DISH-TESTS. 285 

a varnish, the carbon-black and other pigments being ground in 
practically a linseed-oil varnish, and are comparable with No. 1, to 
which they are closely related. The dish painted with only one 
coat showed a little deterioration at the end of the fourteenth evapo- 
ration, while the dishes painted two coats were uninjured, the coating 
being as elastic and tough as when first applied. 

No. 4. Iron-oxide paints consisting of more or less iron oxide 
with more or less silicious matter, and compounds of lime and mag- 
nesia. They were of different grades and qualities, were as a rule 
well ground and spread well. Under test on the dishes painted with 
one coat, after the fifth exposure many rust-spots appeared. Those 
painted two coats were refilled six times, and on them the rust was 
plainly discernible to the eye. 

No. 5. Graphite paints and silica-graphite compounds. These 
paints were received from the several manufacturers in the form 
of a stiff paste, and when mixed, ready to apply, 4^ parts of paste 
to 3i parts, by weight, of boiled linseed-oil were used. The dishes 
painted with one coat were evaporated ten times. After the fifth 
evaporation a few specks of rust were noticeable, and the number 
gradually increased after each successive evaporation. After the 
tenth exposure some slight difference between them was noticeable, 
but not much. The dishes painted two coats were exposed thirteen 
times in two years, and none of them showed any rust or indication 
of rust. The natural toughness and elasticity of the paint still 
remained. 

It will be noted that there is a wide discrepancy in the results 
of the dish-test of Mr. Toltz, as above, of the graphite paints, both 
the natural amorphous pigments and the compounded silica-graphite 
pigments, and the plate-test given by Professor Smith of pure flake- 
graphite mixed with raw linseed-oil that gave 215 pounds of corrosion 
to 1500 square yards. This, no doubt, is due to the repellent nature 
of the pure flake-graphite; the pigment does not take kindly to the oil, 
any more than soapstone does. Raw oil, even if pure, contains from 
5 to 7 per cent of water, that renders a combination of the graphite 
and oil quite uncertain unless under the influence of heat. The boiled- 
oil vehicle with pure flake-graphite, used by Professor Spennrath in 
his experiments (hereafter referred to) with paint-skins detached 
from the metal surfaces, withstood an exposure in a pure water-bath 
for six weeks without injury other than a slight loss in weight of the 
skin. Moisture in the oil in this case was eliminated, as in the case 



286 FERRIC-PAINT TESTS. 

of Mr. Toltz's graphite paints, and the merits of boiled oil as a vehicle 
for most paints over raw oil is sustained in these experiments, as it is 
in daily practice elsewhere. 

United Stales Navy-yard Paint Tests. 

The result of these tests corroborate the series of tests made by 
order of the Secretary of the United States Navy in 1884-5.* By 
request, sixty paint firms submitted seventy-five different paints for 
test, which were applied to five hundred test-plates, and then im- 
mersed in sea-water at foiu" navy-yards, and upon one government 
vessel in service. The paints that successfully withstood the test 
and received an order of merit, were red lead, zinc oxide, carbon, and 
graphite compounds. The so-called asphaltum paints were at the 
bottom of the list in the no-merit column. Evidently there has been 
slight improvement, if any, in this class of paints since the date of 
the U. S. Navy tests to the present time, and one can but wonder, 
in the face of repeated and recorded failures, that they ever receive 
an application to a ferric structure, ashore or afloat. Lead, zinc, 
carbon, and graphite compounds maintain their supremacy for gov- 
ernment work. In other tests of commercial and special paints, 
where the tests have been carried to the destruction of the coating 
as a whole, the partial destruction of the vehicle was generally fol- 
lowed by the disintegration of the weaker substances comprising 
the pigment, such as the carbonate and sulphate of lime, asphaltum, 
iron oxide, and the various color pigments, viz., the ochres, umbers, 
blues, greens, carmines, yellows, etc. The only pigments practically 
unaffected by the destructive element were the graphites, the silica, 
barytes, slag, slate, and brickdust. Other adulterants were but little 
affected, some of them being partly recoverable, which was also the 
case with the red lead, white lead, and zinc-oxide pigments. 

Commercial Coal-tar Paints. 

Fourteen commercial paints, principally of the coal-tar and asphalt 
class, were tested under uniform conditions, viz: Wrought-iron 
plates, free from mill-scale, were coated with two coats each of the 
following paints. All of the coatings were perfectly dry before the 
second coat was applied. When the second coats were dry and 

♦Transactions American Society Mechanical Engineers, 1894, Vol. XVI, 
Paper No. 625, pp. 399-402. 



FERRIC-PAINT TESTS. 287 

hard; one set of the plates was immersed for two months in sea-water, 
and another set was exposed to atmospheric influences, where com- 
bustion gases from locomotives and coke ovens reached them freely 
at all times, this exposure being similar in all respects to that of rail- 
way-bridge paints.* 

No. 1. Carbonizing coating (Cohen Mfg. Co.). 

Physical properties, — Very black, good body, spread well, and 
covered a large surface, coating smooth. 

Drying properties. — Poor. 

After 24 hours, wet. 
" 48 '' quite wet. 
''72 '' slightly wet. 
'' 96 '' dry. 

Physical test. — In sea-water, much rusted and blistered, paint 
easily rubbed off, and in bad condition. Atmospheric exposure, 
condition of coating, fair. 



No. 2. Durable metal coating (Edward Smith & Co.). 
Physical properties. — BrowTiish color, thin and required a large 
quantity to cover; adhered poorly; coating thin and uneven. 
Composition. — Linseed-oil, asphaltum, and kauri-gum 
Drying qualities. — Dried slowly. 

After 24 hours, wet. 

''48 " tachy. 

"72 " slightly tachy. 

" 96 " dry. 
Second coat dry in 7 da3rs. 

Physical tests. — ^In sea-water, much rusted, peeled off easily, 
and blistered. 

Atmospheric exposure. — ^Rusted badly on edges of the plate. 



No. 3. Turpentine asphaUum (C. A. Reeves & Co.). 
Physical properties. — Coating thick and uneven; required a 
large amount to cover. Spread poorly ; adhered well. 



♦Transactions American Society Mechanical Engineers. Experiments by 
J. H. Pennock, Chemist. • Vol. XXIT, 1900, 1901, Paper No. 901. 



288 FERRIC'PA INT_ TESTS. 

Drying properties. — 



After 24 hours, quite wet. 
''48 '' tachy. 
''72 " almost dry. 
" 84 " dry. 



Physical test, — In sea-water, much rusted, paint peeled off in 
spots; did not rub off as easily as No. 2. In two spots corrosion 
had eaten through the plate. 

Atmospheric exposure. — Condition very bad, rusted all over. 



No. 4. "5." Black varnish (Mica Roofing Co.). 
Composition, — ^A coal-tar product, contained light oils, naphtha- 
lene and anthracene, not indicating much pitch. 

Physical properties. — Very fluid, gave a smooth coating. 
Drying properties, — 

After 24 hours, wet. 
"48 " quite wet. 
"72 " sUghtlywet. 
"96 " dry. 

Physical test. — In sea-water, fairly free from rust, but coating 
was rough and uneven, broken off in many places. 

Atmospheric exposure, — ^Rusted badly along the edges. 



No. 5. AsphaUum paint (C. E. Mills & Co.). 

Composition, — ^Asphaltiun, petroleum, and some linseed-oil. 

Physical properties, — ^Thickens on exposure to air; gives a thick 
uneven coat. 

Drying properties. — Dry in 66 hours. 

Physical test. — In sea-water, much rusted and blistered, large 
part of the plate coating entirely gone. 

Atmospheric exposure, — Plate in very bad condition. 



No. 6. Black diamond paint (C. W. Reeves & Co.). 
Composition. — Pitch and dead oil. 

Physical properties. — Quite fluid, spread well and adhered well 
and gave a good even coating; smells of coal-tar.^ 



FERRIC-PAINT TESTS. 289 

Drying properties. — 

After 24 hours, slightly tachy. 
''36 " ahnostdry. 

" 60 '' dry. 

Physical test, — ^In sea-water, did not blister, rusted considerably, 
paint off in spots. Not so good condition as No. 4, but better than 
Nos. 1, 2, 3, and 5. 



No. 7. "A," Varnish (Mica Roofing Co.). 
Composition. — ^Asphaltum, in petroleum spirits. 
Physical properties. — ^Fluid, spreads well and adheres well; gave 
a thick coating fairly smooth, smells strongly of petroleum spirits. 
Drying properties. — 

After 24 hours, quite wet. 
"48 " slightly tachy. 
'' 60 '* dry. 

Physical test. — ^In sea-water, much rusted, not blistered, paint off 
in numerous spots. In bad condition. 
Atmospheric exposure. — Badly rusted. 



No. 8. Mineral rubber (Assyrian Asphalt CJo.). 

Physical praperties. — ^This paint was so thick and viscous that 
it could not be applied without thinning. When thinned with naphtha 
it did not work satisfactorily, and the experiments with it were aban- 
doned. (See Chapter XII.) 



No. 9. Black roofing paint (Samuel Cabot). 

Composition. — ^Pitch dissolved in light petroleum oil. 

Physical properties. — ^Fluid, gave a smooth coating that adhered 
well. Smelt of tar-oil. 

Drying properties. — ^Dried in 56 hours. 

Physical tests. — In sea-water, rusted badly, coating off in many 
spots, rubbed off easily. 

Atmospheric exposure. — Rusted badly all over. 



290 FERRIC-PAINT TESTS. 

No. 10. Black paint (Thomas Mfg. Co.). 

Composition, — A coal-tar paint with a heavy oil menstruum. 

Physical properties, — Much like No. 9. 

Drying properties, — 

After 24 hours, wet. 

48 ' ' tachy. 

60 ' ' dry. 

Physical tests. — In sea-water, paint came off easily; many rust- 
spots. 

Atmospheric exposure. — Condition fairly good. 



it 



No. 11. Slag cement paint (Barrett Mfg. Co.). 

Physical properties, — ^A coal-tar paint, producing a coating similar 
to Nos. 9 and 10. 

Drying properties, — Dry in 60 hours. 

Physical tests. — In sea-water, had a tendency to peel ofif; some 
rust-spots noticed. 

Atmospheric exposure, — Plate badly rusted. 



No. 12. ''Ferrodor'' (Wm. Somerville's Sons). 

Composition. — Graphite, turpentine, oxide of iron, and linseed-oil. 
A compound or patent paint. 

Physical properties, — Color, purplish gray; coating, very thin; 
three coats recommended by the manufacturers. 

Drying properties. — Dry after 48 hours. 

Physical tests. — In sea-water, paint peeled ofif badly, plates very 
much corroded, bad condition generally. The graphite settled to 
the bottom of the can in a tenacious pasty mass, and the paint was 
spread with great difficulty. 



No. 13. ^'Anioxide.'' Ready mixed paint (Harrison Bros.). 

Physical properties, — Bright-red color due to red lead; very 
fluid; spread well, giving a smooth even coating that adhered well. 

Drying properties. — Dry after 48 hours. 

Physical properties, — In sea-water corroded badly, paint peeled 
off, plate much rusted. 

Atmospheric exposure. — Paint turned black, plate badly rusted 
and scaled oflT. 



FERRIC-PAINT TESTS (BAKER'S). 



291 



No. 14. ''CrysolUe'' (Solvay Process Co.). 

Composition, — A paint made from coal-tar (special process). 

Physical properties, — Deep-black color; spread well and adhered 
well, giving a smooth even coating, rather thick; containe^l 10 per 
cent free carbon. 

Drying properties, — Dry in 36 hours. 

Physical tests, — In sea-water no corrosion or blistering; had a 
slight tendency to peel. 

Atmospheric exposure, — Plate slightly rusted; stood the action 
of combustion gases better than any of the other competitive paints. 
There was no tendency of the paint to run or crawl when applied to 
any metallic surface at ordinary temperature. (See Water-pipe 
Coatings, Chapter XII.) 

A number of commercial ferric paints were tested (1899) by 
Prof. Ira O. Baker for their comparative resistance to heat, sea- 
water, strain, elasticity, the fumes of sulphuric and nitric acids 
also carbonic acid, with the following results:* 

The samples of paint named were furnished by the manufacturers 
for the purpose and mixed with linseed-oil as directed by them, and 
spread on clean bright wrought-iron test-plates. 



Reference 
Number. 


Kind of Paint. 


Weight per 
Gfllion. 


How Received from the Maker. 


1 


Red lead 


PoundB. 

31.72 

21.24 

11.34 

14.13 

12.99 

9.49 

8.57 

8.84 

9.52 

13.09 

9.61 

17.06 


Dry pigment. 
Paste. 


2 


White lead 


3 


Purple iron oxide 


Mixed ready for use. 
Dry pigment 
Mixed ready for use. 

tf H H it 

Dry pigment. 
Mixed ready for uae. 


4 
5 
6 

7 


Chattanooga iron oxide 

William sport iron oxide 

Detroit superior graphite. . . . 
Mexican jrraDhite 


8 


Dixon's graphite 


9 


Trinidad asphalt. 


If tf ft It 


10 


Bessemer paint. 


U If II u 


11 


Carbonizing coatine. 


II It It It 


12 


LithcMren silicate 


Paste. 









A sample of each paint in one and two-coat work was exposed 
to heat and the products of combustion in the smoke-flue from a 
boiler burning bitumnous coal. The condition of the plates on 
removal was: 

1. Red lead. — ^Entirely dead and very brittle. The powdery resi- 
due was easily removed, bringing into view the base metal. 

* Railroad GazetU (New York), March 10, 1899. 



292 FERRIC-PAINT TESTS (BAKER*S) 

2. White lead. — ^Blistered and baked. Coating was brittle and 
entirely dead, and very easily removed. Removal of the residue 
exposed the base metal. 

3. Purple iron oxide. — Covered with blisters. Paint soft and 
easily removed. Removal of the bUsters exposed the base metal only 
to a slight extent. 

4. Chattanooga iron oxide. — Considerably blistered, moderately 
soft, adhering well. Removing the blisters did not expose the base 
metal. 

5. Williamsport iron oxide. — Covered with blisters. Paint tena- 
cious, adhering well. Removing the blisters exposed the base metal. 

6. Superior graphite. — Small blisters. Paint hard and difficult 
to remove. Almost impossible to expose the base metal. 

7. Mexican graphite. — Soft and readily removed. Removal of 
blisters did not expose the base metal. 

8. Dixon's graphite. — Very few blisters. Paint hard and very 
adherent. 

9. Trinidad asphalt. — Smooth; considerable of the paint melted 
and ran off. The portion remaining was hard and brittle. 

10. Bessemer paint. — Soft and hard to scratch off. 

11. Carbonizing coating. — Ridges very conspicuous. Paint finn 
and adherent. Removal of blisters exhibited a very porous, loose 
structure of the paint. 

12. Lithogen silicate (white paint). — Substantially the same as 
the white lead. 

Effect of salt water. — ^A set of the plates were exposed to a satu- 
rated solution of sea-salt (brine) for seven weeks, the plates being 
frequently withdrawn and allowed to dry. Their condition at the 
end of the exposure was: 

1. Red lead. — Hard and adhering. Metal clean and bright under 
the two coats, but badly rusted imder the single coat. 

2. White lead. — ^Hard and adhering. Clean metal under two 
coats, rust under the single coat. 

3. Purple iron oxide. — Firm and adhering. Occasional rust-spots 
throughout. 

4. Chattanooga iron oxide. — ^Firm and adherent. No rust any- 
where. 

5. Williamsport iron oxide. — ^Firm, and adhered only fairly well. 
Metal under two coats bright, but under the single coat considerable 
rust. 



FERRIC-PAINT TESTS (BAKER'S). 293 

6. Detroit superior graphite. — ^Hard and adhering well. Rust 
under the single coat, none under two coats. 

7. Mexican graphite. — Elastic and easily removed. No rust. 

8. Dixon's graphite. — Grood condition. Elasticity only slightly 
impaired. Easily removed no rust. 

9. Trinidad asphalt. — Seemingly unaffected. No rust anywhere. 

10. Bessemer paint. — ^Moderately hard, peeled off easily, no rust. 

11. Carbonizing coating. — Peeled off easily. Rust beneath both 
the one and two coats. 

12. Lithogen silicate. — Hard and adherent. No rust-spots. 

No report was made of the condition of the paints exposed to 
atmospheric influences, evidently for the reason that the exposure 
period had not been long enough to materially affect any of the 
paints when the test closed. 

A set of the plates were exposed to the fumes of strong sulphuric-, 
nitric-, and carbonic-acid gases for five weeks. The action of the sul- 
phurous gas was characterized by its bleaching power upon the paints 
and the disintegrating effect on the iron under the paint, also the 
formation over the entire area of blisters, under which was found a 
moderately hard whitish deposit. 

The order of merit for the several paints was: Trinidad asphaltum, 
Carbonizing coating, Dixon's graphite, Red lead, Lithogen silicate, 
Mexican graphite. Purple iron oxide, Superior graphite, Bessemer 
paint, Chattanooga iron oxide, Williamsport iron oxide. White lead. 

The effect of the nitric-acid gas was substantially the same as the 
sulphurous gas, except that the single coatings of the paints were 
completely destroyed. The order of merit for the double coatings 
was: Trinidad asphaltum, Lithogen silicate. Red lead, White lead, 
Purple iron oxide, Superior graphite, Mexican graphite, Chattanooga 
iron oxide, Williamsport iron oxide, Dixon graphite. Carbonizing coat- 
ing, Bessemer paint. 

The carbonic-acid gas in large quantities, supplemented by mois- 
ture, had only an almost imperceptible effect upon any of the paints. 

To test whether any of the paints would crack during the elonga- 
tion of a painted bar, strips of machine steel 2''Xi''Xl8" were painted 
two coats, and after drying for two months were submitted to a strain 
of 16,000 pounds per square inch; the paint in every case remained 
firm and close-adhering. The stresses were then increased slowly 
beyond the elastic limit of the steel, and in no case did the paint 
crack before that point was reached. 



294 FERRIC'PAINT TESTS (BAKER'S). 

It was noticed that after passing the elastic limit of the sted, 
the paints were marked by a series of lines arranged in herring- 
bone patterns, that were alike on both sides of the bar and alike 
situated. Evidently the lines were due to a rearrangement of the 
atoms of the steel bars w^hile under strain, the centre atoms moving 
less freely than those near the comers and edges of the bars, the 
paint naturally following the particles of steel that they covered. 
Naturally the heavier coatings of the paints were the least elastic. 
The order of merit in the elasticity test was: Trinidad asphaltum. 
Carbonizing coating. Purple iron oxide, Dixon graphite, Mexican 
graphite, Superior graphite, Williamsport iron oxide, Chattanooga 
iron oxide, Bessemer paint, White lead, Red lead, Lithogen silicate. 



CHAPTER XXX. 



PAINT TESTS ON RAILWAYS. 



New York Elevated BaUways. 

The physical condition and the extent of corrosion on all parts 
of these structures have been freely commented upon by the technical 
engineering journals and the daily press. When originally erected, 
no attempt was made to remove the mill-scale, and none has been 
made since, presumably because of the impossibility of its success, 
and the cost. The composition of the paint which has been repeatedly 
applied to them has been kept very uniformly good in quality, but its 
application has been solely for appearance, as no paint can now reach 
the seat of corrosion underlying all the coats of scale, dust, cinders, 
and paint. 

The renewal of the whole structure will probably be necessary in 
less than a hundred years from its erection. 

The composition of the paint used is given by the chief engineer 
as follows: 

For Fifty Gallons of Paint, Olive-drab Color. 

Summer Winter 

Formula. Formula. 

White lead, Jewett's best 300 pounds 275 pounds 

Bridgeport zinc oxide, strictly best quality 175 " 150 " 

French ochre " " " 100 " 90 

Prussian blue " " " 1 " 1 " 

Lampblack " " " i " J " 

576i pounds 516i pounds 

The above pigments are ground in Campbell & 
Thayer's raw linseed-oil. The weights given include 
the necessary oil to grind the pigments to a paste. 

When applied, it was mixed with 

Boiled linseed -oil, Campbell & Thayer's 8 galL 9 galL 

Raw linseed-oil, " " ** 15 " 15 " 

Spirits of turpentine, first quality 3 " 3 " 

Liquid or japan drier, " " 2 " 3 " 

28 galL 30 galL 

295 



296 PAINT TESTS ON RAILWAY STRUCTURES. 



New York Elevated Railway Viaduct, 

The viaduct over the Harlem Station of the New York Elevated 
Railway at 155th Street was oil-coated, and received iron-oxide 
paint coatings at the time of its erection, and within five years of its 
eompletion had developed corrosion to such an extent that in 1897 
the sand-blast was used to clean it preparatory for another effort for 
its preservation. This sand-blast process cost about $10,000, or over 
fifteen cents per square foot to apply, or about seven times more than a 
properly selected method of procedure and paint would have cost in 
the first place, and then only the lower and accessible sides or parts 
in sight received treatment. About 50,000 square feet of surface was 
cleaned by the sand-blast, to the bright iron, removing about 12 tons 
of old paint, scales of rust, and cinders, showing a unmber of distinct 
layers of highly corroded matter. 

Seventeen panels of lattice-truss, floor-beam and buckle-plates, 
supporting the paved carriage roadway and footpaths overhead, 
about 2825 square feet of surface each, and numbered consecutively 
1 to 17, were then painted with the same number of selected protec- 
tive coatings furnished by a like number of paint firms in competition 
with each other. The several coatings were applied in strict con- 
formity to the directions received with each brand of paint, the appli- 
cation being to the bright iron as left by the action of the sand-blast, 
.and within 3 to 4 hours from the time the sand-blast ceased action. 
Every possible condition was brought into bearing to make the test 
one of a practical and commercial nature as well as of .scientific value, 
absolutely without prejudice or favor in any respect. From the 
prominence of the structure in an engineering view, and its situation 
exposed to storms, sea air, fog, cinders, steam, and gases from scores 
of locomotives in constant service beneath it, nearly all of the metal 
being within a few feet of the tops of engine-stacks and receiving the 
products of combustion under blast-action and in an approximately 
closed space, the future result was anxiously looked for as an impor- 
tant demonstration of the practical value of the several best pro- 
tective coatings in the market. 

After an exposure of about nine months, and while a few of the 
coatings, viewed from the station platform, showed slight evidences 
of failure, a thorough examination of the condition of each panel 
was made by a prominent civil engineer of New York City. This 



PAINT TESTS ON RAILWAY VIADUCTS. 



297 



report is of extreme interest, and is summarized, viz., 100 rating as a 
perfect condition of the coating.* 



5z; 



o o 

B O c3 



Kind or Name of Paint. 



1 


3 


2 


2 


3 


2 


4 


2 


5 


3 


6 


2 


7 


2 


8 


2 


9 


2 


10 


4 


11 


2 


12 


2 


13 


2 


14 


2 


15 


2 


16 


2 


17 


2 



Rate of 
Drying. 



Lead, graphite and lucol-oil 

Amorj^ous graphite, Detroit Co. — L.S.G. 

Red lead, antoxide, F. and P 

Graphite (kind not stated) 

Nobrac (trade mark) 

Carbon Black (F. W. Devoe & CJo.) 

Durable Metal Coaling (£.Smith'8 varnish) 

Black Manganese (iron paint) 

Carbonizing Coating (trade mark) 

Mineral Rubber (no particulars) 

Black varnish (composition not given). . . 

Carbon paint (no particulars) 

Graphite (Standard Oil Co.). Kind not 
stated .' 

Dixon Co. Silica graphite, mixed paint . . . 

Asphaltum (California Co. brand) 

Ruoerine (trade mark). CJoal-tar compo- 
sition 

Black diamond (trade mark) 



Medium 

Slow 

Fast 

Slow 

Medium 

Slow 

Slow 

Fast 

Slow 

Fast 

Medium 

Medium 

Medium 
Slow 
Very Slow 

Medium 
Medium 



5^ 



I 



p s 



97 
80 
25 
75 
99 
85 
75 
30 
80 
78 
58 
92 

67 
70 
65 

58 
70 



8 21 

K ^ C 

^6 



a 
h 
c 
d 
e 
d 
d 

f 
a 

i 

a 



a 
d 

k 
I 



a. Very little rust. Paint crumbles in places as though rotten. Easily re- 
moved. 

b. Fair condition, but discolored ; rust coming through. 

c. Very badly rusted. 

d. Rusty, but not deep. 

e. Slight rust on top flange of one girder; rest of girder clean. 
/. Rust very deep ; buckle plates bad. 

g. Area of rust-spots small; rust not very deep. 

h. Rust very bad and deep. 

j. Deeply rusted ; buckle plates still good. 

k. Rust very deep and angry; buckle plates mildewed. 

I. Small pimples of nist, as though formed under the paint 

Panel No. 1 was an outside one, and the first to be sand-blasted 
and painted, in some parts with two and in others three coats of 
paint, in the clear hot days of summer, a material advantage in its 
favor. The sand-blast was then shifted to the southern end of the 
viaduct; and panel No. 17, also an outside one, was the next one 
cleaned and painted in hot clear weather, and so on consecutively, in 

* Engineering News, September 23, 1897, Illustrated; Engineering Record, 
September 25, 1897, Illustrated. 



298 PAINT TESTS ON RAILWAY VIADUCTS. 

the reverse order of the panel numbers, back to No. 1 ; panels Nos. 7 
to 2 having been done late in the fall under unfavorable conditions 
as to the spreading and drying of the paint in addition to the other 
objectionable conditions. About 80 square feet of panel surface 
was cleaned per hour, or 600 square feet per working day; each 
panel requiring from five to six days to clean and paint it. 

At the end of about a year the condition of all of the paints was so 
unsatisfactory that the viaduct was repainted without removing, 
only in a perfunctory manner, the old test coatings with their 
fast-forming burdens of rust; and this competitive test came to an 
inglorious end. 

The result could have been foreseen from the first, before a single 
truss or pound of material had been placed in position, or was even 
out of the construction shops, had not commercial greed, official 
indifference or ignorance, either one or all, ruled the matter. 

The destruction of the tubular railway bridge over the St. Law- 
rence River at Montreal, Canada, had not become a fact so musty 
with age as to have escaped attention concerning the dangerous effect 
of hot combustion gases upon any paint coating in a confined space. 
Corrosion history blindly repeated itself when the viaduct material 
was first painted by the contractors, then repeated the "Comedy 
of Errors" when it was erected and again when it was sand-blasted 
for its final fiasco. The plain facts of the painting after the sand 
blast action are, that the coat ngs were destined for an early destruc- 
tion from the beginning, by reason that the first coat was appUed in 
an atmosphere saturated with the hot vapors of combustion and 
steam, which were so corrosive that the freshly cleaned surface of 
the metal showed a blush of rust within an hour after cleaning, and 
if left for three hours the rust could be wiped off by the hand. The 
paints were spread in this atmosphere, and before they could in 
any measure dry, so as to be in any degree resisting, they were thor- 
oughly impregnated by the hot gases and steam which left their con- 
densed strength upon the surfaces of the green paints. The second 
and subsequent coats were not only applied under the same atmos- 
pheric conditions as to the hot vapors and cinders, but had the 
condensed products of combustion sandwiched between them. 
Probably a baked japan or Bower-Barff coating are the only ones 
which would have successfully met the situation, which is an excep- 
tional one. Such coatings applied at first, would not have cost one- 
half as much as the sand-blast and the several coatings applied in 



PAINT TESTS ON RAILWAY VIADUCTS. 299 

the first and subsequent stages, and would have been thoroughly 
protective and avoided nearly all the future expense in the care of 
the structure so far as the painted surfaces are concerned. 

A paint test of an extended character has been in progress for 
the past few years by Mr. Geo. W. Webster, C.E.,* to determine 
the best paints for use on the city street iron bridges, crossing the 
railroads within the city limits of Philadelphia, Pa. 

Fifty-four sample plates of iron 12"X24" were coated by twenty- 
two manufactures of paint, and exposed at a number of places on 
the street viaducts, in situations that were as nearly uniform for the 
several competitive coatings as possible to provide. The test coat- 
ings were changed in their location as circumstances required to 
equalize the exposures, which were very severe, the clearance between 
the top of the locomotive stacks and the metal work of the bridges 
being only two to three feet. 

The samples of paint submitted included the most prominent 
proprietary paints, including the carbon and graphite classes. 

The results of only a few months' test demonstrated that on the 
lower surfaces of the plates on the bridge structure no paint was 
able to resist the mechanical injury from the sand-blast action of 
the locomotive exhaust. These situations are now protected from 
this action by wooden sheeting a few feet in width on the line of the 
exhaust. 

On the upper side of the test-plates, subject to moisture, com- 
bustion gases, and deposits of cinder, the results were more satisf ac- 
tor}', but were not conclusive as to the relative merits of the samples, 
due to the difficulty in comparison on the basis of truly identical 
conditions. 

The general trend of the results was, the subsequent selection of 
certain of the proprietary paints for a trial on bridges, and component 
parts of bridges, under a general formula, having red lead as the 
principal pigment, viz: 

Red lead, two coats over shop coat of raw linseed-oil for inclosed 
space of structures. 

Red lead over shop coat of raw linseed-oil and two coats of white 
lead three parts, and zinc oxide one part, for the field work. 

Indian red, one coat over shop coat of oil. Red lead, third coat. 

* Chief Engineers Bureau of Surveys, Department of Public Works, City of 
Philadelphia, Pa., 1902. 



300 PAINT TESTS ON RAILWAY VIADUCTS. 

Indian red, one coat over shop coat of oil. White lead three 
parts, zinc oxide one part, for field work. 

The Gray's Ferry deck bridge over the Schuylkill River was in- 
cluded in the test, being painted with the following paints, all applied 
as a first or shop coat: 

Nobrac, lucol-oil paint, rubber paint, Bessemer paint, antoxide, 
durable metal coating, red lead. Above the deck for the second full 
coat, red lead was applied, except in one case, where white lead and 
zinc oxide were used. Below deck, for the third full coat, the same 
paint was used as for the shop coat, except in one case where white 
lead and zinc oxide were used for both the second and third coats. 

The above coatings were applied in 1899-1900, and the time 
since then has been too short to note any material difference in their 
condition. The general practice in painting ferric structures by 
the Board of Public Works of the city of Philadelphia for a number of 
years has been the use of red lead and lampblack for a first or prim- 
ing coat at the shop, followed by two coats of the same paint in dif- 
ferent shades of chocolate color for the field coats, though in some 
cases white lead and zinc oxide have been the field coats. That the 
above paints have not proven satisfactory in the presence of com- 
bustion gases and other influences incident to their location is evi- 
dent from the above experiments to correct their deficiencies. In 
connection with the same matter, it may be of interest to note that 
the train-shed roof of the Broad Street Station of the Pennsylvania 
Railroad at Philadelphia, which was painted with red lead and lamp- 
black, is seriously affected by the corrosion of the roof-trusses. This 
structure has had extremely good care since its erection; but corro- 
sion has established itself, owing to the early decay of the red-lead 
coatings, and will soon require cleaning by the sand-blast to correct 
the mistake of using red lead for train-shed painting. (See Chapter 
XXXVI, Changes in Pigments.) 

Inffuences that Affect Paints, 

Some of the influences that affect the life of a paint coating have 
been determined by the experiments of Prof. J. Spennrath,* from 
whose essay the following excerpts are selected. The experiments 
were made upon paint-skins alone, not upon a painted surface. The 
skins were made from chemically pure, finely ground flake graphite 

* "Protective Coverings for Iron." Railroad Journal, New York, 1896. 



PAINT TESTS. SPENNRATWS EXPERIMENTS. 301 

and linseed oil, applied to zinc plates in two coats, each of which 
was allowed to harden thoroughly. The plates were then placed in a 
dilute solution of sulphuric acid and the zinc dissolved. The paint- 
skins were then used for testing by immersion and exposure for 
six months in a number of liquids and gases, as follows : 

Immersion Tests, 

In pure rwn-water the skin remained cohesive, 
even elastic, was of dull color, noticeably injured, and 
lost in weight 10.4 per cent. 

In sea water the skin remained uninjured in tex- 
ture and lustre, with a small loss in elasticity, and in 
weight 4.52 " " 

In a 10 per cent solution of common salt the 
skin was but little affected in lustre and elasticity, 
but lost in weight 2.4 " " 

In a 10 per cent solution of sal-ammoniac the 
skin was unchanged. Lost in weight 3.5 " " 

In a 5 per cent solution of sulphuric acid the skin 
remained unchanged. Lost in weight 1.65 " " 

In a twenty-four-hour immersion in hot water 
160® to 170° F. the skin was materially affected in 
texture and color. Lost in weight 9.83 " ** 

In an aqueous solution (alkaline) of mineral-coal 
ashes the skin was materially affected. Lost in 
weight 14.8 " " 

In a 1 per cent solution of soda the skin after 
three days was vividly affected, and after a few days, 
more exposure was destroyed. 

In a 5 per cent solution of nitric acid the skin 
was destroyed. 

In a 10 per cent solution of the chloride of mag- 
nesium the skin was unchanged, but lost in weight 1.1 " " 

Exposure Tests in Closed Vessels. 

Over sea-water for six months the skin was unin- 
jured in color or texture, but had become somewhat 
viscous; no loss in weight. 

Over dry chloride of calcium (anhydrous) the skin 
was not at all affected, and gained in weight 0.46 " " 



302 PAINT TESTS, SPENNRATH'S EXPERIMENTS. 

Over acetic acid, fuming muriatic acid, nitric acid, 
ammoniacal liquor, liquid sulphate of ammonium, 
a solution of gaseous sulphurous acid and water, all 
the skins were destroyed in a few days. 

A skin made from red lead and linseed oil, exposed 
for forty-eight hours to an atmosphere of hydric 
sulphide, became black and rough, dull in lustre, and 
increased in weight 1.5 per cent. 

The changes here indicated relate solely to the vehicle, as the 
graphite pigment was passive to the action of any of the destruc- 
tive agents. Wherever the skins were destroyed every other oil- 
paint coating would have been likewise destroyed, whatever pig- 
ment was in it. In the other cases where changes in the vehicle 
are noted, the change of the skin appears to be wholly imlike that 
which would have occurred had it been attached to any surface. 
The professor has evidently found this to be the case, judging from 
some of his notes preceding the record of his tests. There are 
commercial paints — notably those made from amorphous mineral 
graphite, that containing less graphitic carbon, and combined with 
silica and a small amount of mineral oxides — that would have afforded 
a better protection to the vehicle than the chemically prepared 
graphite used in these experiments, the cost of which would prob- 
ably bar it from forming any part of a protective covering for iron. 

Spennrath^s Temperature Tests. 

A number of graphite paint-skins of the same character as those 
used in the above immersion and exposure tests; also, seme three- 
coated skins made with other pigments and linseed oil and mixed 
with turpentine and other driers, and mineral oil, were submitted 
to constant temperatures of 122°, 203°, and 248° F. for five days. 
Briefly the results were: 

All the skins shortened from 1.2 to 4.3 per cent, averaging 3.76 
per cent, and lost in weight from 2.11 to 8.3 per cent, averaging 
5.82 per cent; the greatest change occurring in the single instance 
of a graphite skin exposed for five days to a temperature of 248° F., 
which shortened 5 per cent and lost in weight 9 per cent. 

The smallest change was also in a graphite skin exposed for five 
days to a temperature of 120° F., that shortened 1.2 per cent and 
lost in weight 4.4 per cent, and, though visibly affected in elasticity. 



PAINT TESTS. SPENNRATH'S EXPERIMENTS. 303 

was changed the least in other respects. All the other skins became 
brittle and stiff, broke easily when bent sharply, and were darkened 
in color. The white-lead skin changed to a faint yellow, the sul- 
phide of lead and zinc oxide skin to an intense yellow. 

The addition to the linseed oil of 10 per cent of either the oil of 
turpentine or other driers, or a mineral oil, or other fatty non-drying 
oils, including some gum copal, had no effect whatever to resist the 
changes effected by the heat. Not more than 10 per cent of mineral 
oil could be added to the Unseed oil, as it rendered the paint \dscous 
after drying. The Bessemer paint-skin, the pigment being a ground 
furnace slag, and the linseed-oil vehicle having an addition of a non- 
dr^dng fatty oil witTi some gum copal, was just as sensitive to the 
heat as any oil paint. 

Generally the action of the heat was less marked upon the 
graphite skins, which were less brittle than those made from white 
lead or zinc-white. The red-lead skins were especially sensitive to 
mechanical influences. 

These changes are easily accounted for. The oily, repellent 
nature of the flake graphite prevented it from bonding to the oil 
vehicle as firmly as the other natural pigments and those of higher 
specific gravity will do in both a green or a thoroughly dried paint. 
While it is generally known that heat is destructive to oil-paint 
coatings, it does not follow that the coatings are so sensitive to its 
action that it may be deemed the principal cause of their failure. 
The engine and fire-rooms of ocean steamers and war-vessels are 
exposed to temperatures of 120° to 140° F. for months at a time, 
and many times in succession in atmospheres heavily charged with 
moisture and other vapor, without any material disturbance to the 
protective character of the coatings. 

There are commercial paints in extensive use, subject to tem- 
peratures of 300° to 350° F. under pressures of steam, which preserve 
their integrity after years of exposure. In these instances the life 
of the coating depends quite as much upon the vehicle as upon the 
pigment. 



CHAPTER XXXI. 

PAINTING BY SPRAY. 

PAiNTma by spray or the air-brush has lately come largely into 
use; in fact, it would have been impossible to have covered, in any 
acceptable manner, the World's Fair Buildings erected since 1890, 
without the use of the paint-spray process. 

At the Columbian Exposition, the results of the spray method 
of painting, compared with the use of the hand-brush, were: A 
corps of hand-brush painters, working in the usual manner of apply- 
ing kalsomine, averaged about 800 square feet of surface daily, while 
16,000 to 20,000 square feet were covered by a spray-machine in 
eight hours, 30,000 square feet having been reached imder favorable 
conditions. In the Manufacturers' Building, with a daily average 
of thirty men using spray-machines, at the end of eighteen working 
days, 1,332,700 square feet of surface had been covered; equal to 
an average of 2368 square feet per day per man. This was during 
the coldest days of winter, when the water paint, in attempting to 
spread it by hand, froze solid. The spray required about twenty-one 
gallons of kalsomine against twenty gallons by the brush, but the 
saving in labor was nearly twenty to one in favor of the spray process. 

Gas-holders in duty are exceptionally hard to paint. One painted 
by spray, using iron-oxide paint, averaged from 2700 to 2900 square 
feet of surface per hour for three men using two sprays. The amount 
of paint used was not notably more than with the brush. The cost 
of the labor was less than J cent per square yard. 

In the Michigan Engineers' Manual for 1897, Mr. J. J. Huber 
describes an extemporized spray-machine, and the results in paint- 
ing 100,000 square feet of rough hemlock siding with iron-oxide and 
raw-oil paint. 

The contractor's bid for labor, ladders, and brushes, 

the company to furnish the paint, was 35 cents per 100 square feet 

The company to furnish all the material 28 " " " " " 

A lump bid for the labor alone. 30 " " " " " 

304 



PAINTING BY SPRAY. 306 

The mill company built a spray-machine for twenty dollars, and 
the result of its use was, that one gallon of paint covered 150 square 
feet of rough-board surface. Two men covered 5000 square feet 
per day. Coat of the paint applied, ten cents per 100 square feet. 
GoBt of the paint, labor, and apparatus, fifteen cents per 100 square 
feet, or less than one-half that of painting by hand. The paint could 



Fto. 41.— Held spray apparatus at work. 

be applied from eight to ten feet above the spraymens' heads, and 
ordinary laborers could do the work. 

In some experiments m&de by the P. & L. E. R. R,, using a spray 
apparatus for painting box freipht^cars, the time required was thirty 
minutes per car; one man with an eight-inch brush following the 
spray thirty minutes more, or total of one hour per car for each coat. 
To paint a 60,000-pound-capacity coal-car required two men twenty 
minutes each, spraying the lettering not included. 

At the Master Car and Locomotive Painters' Association, in 1897, 



i(06 PAINTING BY SPRAY. 

Mr, H. G. MacMasters, M.C.P.I.C.RR., rei>orted the comparative 
time and coBta in detail of painting box f reigh1>cars by brush and spray. 

With the Brush. With the Spray. 

W(Hk OmML Tims. C«t. Tims. Out. 

Sills, one coat. Mmin. $0.05 13 min. $0.03i 

Edge board, one coat 40 " 0.10 17 " O.Oli 

Body, three coats. 7 hra. 1 .05 84 " 0.21 

Puttying up 1 " . 15 1 hr. 0.15 

Roof, two coate. 30 min. 0.07) 12 min. 0.03 

Trucks, <Hie coat. 1 hr. 0.15 20 '* 0.05 

Blackiiig ironworic 25 min. 0.06} 25 " 0.06t 

Totals. 10 hn. 55 min. $1,631 3 his. 51 min. $0.57} 

Result 1.98 to one in time, and 2.82 in cost, in favor of the spray. 

The danger to the health of the painters in the use of the spray 
is very marked over that in the use of the brush, whether kalsomine, 
iron oxides, or mixed paints are used. In the spraying of lead paints 
or those containing any metallic oxides the dangerous effects are 
greatly increased, even when the greatest potisible care b losed to 
guard against them. The fine mbt-like spray b readily taken 
into the lungs at every respiration, and is more thoroughly intro- 
duced into the system than is possible by absorption from contact in 
painting by hand. 



Fio. 42. — Mathewson's patent helmet for painting by spray or cleaning by 

the sand-blast. 

The extra amount of paint, about 5 per cent, used by the spray 
is offset many times by the saving in labor, as above not«d. 



P.1/.V77.VG BY SPll.lY. 307 

These were applications from compressed-air installations used 
for other purposes thaa painting. 

The merit of oil-paint spray coatings has not been fully estab- 
lished. The spray necessarily carries a part of the air with the con- 
densed moisture in it into the paint, and its subsequent escape by 
expansion and evaporation must result in a more porous coating 
than with paint applied by a hand-bnish. Following the spray 
immediateiy with a brush will remove the porosity to some extent. 
The brushing out of any paint is a great factor in its durability, and 
as the use of the spray renders the employment of a cheaper grade 
of labor more feasible than with the use of the brush, the eflects 
of an indifferent use of paintor'3"elbow-grease" will soon reveal itself 
in the decay of the coating. 



Fia. 43.— Barrel and hand-power spray apparatus. 

Upon metallic surfaces the sprayed paint proves more perish- 
able than when applied to wood. 

For use in the field for repainting iron bridges, walls, and other 
structures there may be a saving in time of the painters, scaffold- 
ing, etc, but the work cannot be as well done as by the use of the 
brush. The cheaper and less responsible laborer generally em- 



308 PMSTING BY SFRAY. 

ployed to use the spray apparatus will neglect the necessary scrap- 
ing and steel-wire brushing requisite in all repainting work, as well 
as the subsequent brushing out of the spray coating, particularly 
in the places difficult of access, where a thorough application of 
the paint is most necesaarj-. Peeling of the paint and corrosion 
promptly follow any extended application of an oil-pmnt spray 
coating on a ferric body in situ. 

The methods of painting were recently discussed by the Western 
Association of Rfdlway Superintendents of Bridges and Buildings, 
in answer to a circular asking for information upon the subject. 
Eighteen answers were received. All were in favor of the sand- 
blast for cleaning either new or old metallic surfaces preparatory 
to painting. 

Six were in favor of the air-spray for some classes of work, three 
were opposed to it, and nine were non-committal. Two who had 
tried it were opposed to it. One superintendent said: "On iron 
bridges other than on plate girders I found that there was more 
paint wasted than applied to the structure. The waste in attempt- 
ing to paint lattice-truss work was very marked, and the coating 
was not equally or well spread." He favored the use of a stiff hard 
brush to insure a close contact of the paint, which he could not 
get with the spray. Too much air was incorporated with the 
paint by the apray, and would not release tself in the drying of the 
paint. It left the coating more porous than with the use of hand- 
brushes. Following the spray with a hand-brush did not materially 
help the coating in durability, when compared with surfaces spread 
on the same structure at the same time, by the same painters using 
hand-brushes, and the same paint. 

"The use of the spray, following it with a heavy hand-brush, 
was admisttble upon some of the large wooden bidklings, as the 
f^Iure of the paint in these cases was not 
attended by corrosion, blistering being the 
principal cause of failure in the sprayed coat- 
ing on wooden and masonry surfaces." 

Generally, heavy or very thick oil paints 
cannot be successfully spread by spray, 
Tn. 44.— Hand apray ""^^ ""^^^ ^^ pressures of sixty or more 
app&rstus. pounds per square inch. This renders the 

use of the spray for oil paint useless, unless power other than hand- 
power is available, or the paint is applied quite hot to im 



PAINTING BY SPRAY 309 

fluidity. If the paint is thinned with benzine or turpentine to the 
point where a moderate pressure of air will enable the spray to 
work without choking in the nozzle, all of the objections to this class 
of paints are increased, as the extra amount of volatiles in them to 
be evaporated leaves the coating more porous and hastens its decay. 
Two coats of such air-sprayed paints are required to equal in pro- 
tecting power one coat of heavy hand-brush work. 

With kalsomining or water paints the spray apparatus finds an 
almost uncontested field, and from the great saving in labor is recom- 
mended. 



CHAPTER XXXII. 



MIXED PAINTS. 



Reputable manufacturers of standard pigments are greatly 
at the mercy of many of the proprietary or patent paints ready 
for use that are a feature of the paint market. 

Standard pigments subsequently appear in these compound paints 
mixed with a variety of well-knowTi inferior substances that by 
some alleged special mechanical manipulation '* developed in our fac- 
tory " makes of them a preeminently superior product, '* wholly 
unlike that produced by the antiquated process employed before the 
advent of our new idea." 

Mixed paints for the great bulk of the paint trade are a con- 
venience that cannot be ignored, and are in conformity with the 
general advancement of the times. Responsible manufacturers 
and dealers in paints furnish them; and they are more uniform in 
quality and color, of better composition, as well as cheaper, than when 
mixed by the individual painter. 

Responsible paint manufacturers inspect and test the quality 
of all their materials, and are certain that thev are standard in all 
respects, more than it is possible for the individual painter to do^ 
however much he may desire to produce a good paint. 

Railway companies and bridge-manufacturing firms, from the 
magnitude of their painted work, are able to employ the necessary 
staff to secure good materials, also the technical knowledge to mix 
and apply them. The application of the paint in a great measure is 
under their control, and the composition of it can be varied to meet 
all the conditions as they arise, and a direct responsibility established 
for any failure in the coating. 

Failures are possible and occur with the best of paints. The 
hardest to locate and most annoying is where the painter, unmindful 
of the atmospheric conditions or the state of the surface he is at work 

310 



MIXED PAINTS 311 

Upon, finding that the paint does not work or spread well, doses it 
with benzine, turps, or some other volatile or vehicle. He does not 
always know or care what the composition of either the paint or the 
added element is, so that it enables him to get through with his work 
without an immediate appearance of distress in the coating. 

Special mixed commercial paints generally require a special 
order of procedure in their application, and when these are faithfully 



Fro. 4fi, — Power paint-ouxer. 

carried out will usually give better results than when the applica- 
tion is left to the ordinary painter's manipulation. 

A few special paints that have pa.ssed the fortuitous require- 
ments of the United States and foreign patent offices, and have 
one or more trade-marks to each combination, are the following: 

Fire- and water-proof. Composition: coal-tar, oil, gypsum, Japan, 
liquid rubber, nitric add, slate-dust, sal-soda, potash, antimony, 
and sodium. 

Another combination of coal-tar, yellow ochre, plumbago, lime, 
salt, and coal-oil. 



312 MIXED PAINTS. 

Fire and acid-proof. Composed of coal-tar, pitch, common 
mineral paint, hydraulic cement, gray ochre, asbestos, slaked lime, 
liquid drier, and litharge. 

In most of the above and in other similar patent compoimds 
the quantities of each substance are not defined, but left to the dis- 
cretion of the user. 

Other compounds of a kindred nature contain saltpeter, sulphur, 
caustic potash, mica, talc, zinc slag, salts of tartar, oxide of copper, 
shellac, sulphate of merciu-y and sulphate of zinc, verdigris, copperas, 
india-rubber, hydrauUc-cement slag, soapstone, solutions of gall- 
nuts, tannin, acetone, yellow soap, lignum vitae, garlic, asafetida, 
and one or more of the list of inert pigments given in Chapter 
XVIII. 

These compounds are recommended as special paints for ferric 
structures. In most cases the merits are so blindly set forth that 
one is in doubt whether it is the preservation or destruction of the 
paint or the covered surface the proprietor wishes to secure. 

As mentioned before, all mixed paints are not necessarily objec- 
tionable compounds, and to be avoided. A combination of pigments 
to secure a desired result is often necessary where the use of one 
pigment would be ineffective. 

In the cases of red lead and lampblack the lampblack delays the 
setting, adds body, prevents the crawl or curdling of straight red- 
lead coatings, and is in every way beneficial to the ph3rsical charac- 
ter of the paint. 

There is enough oxidizing element in the red lead to cause the 
lampblack (which is a slow drier) to dry without usmg a large 
amount of japan or other drier. 

A small amount of French ochre is sometimes added to red- 
lead and lampblack mixtures to give a brighter tone to the choco- 
late color of the mixture, making what is called the "Pullman color." 
This mixture has proved to be very reliable under some severe 

exposures on cars. 

The specifications for the painting of one of the largest bridges 
in the world called for a pure red-lead paint. Analysis of the paint 
after the bridge was painted showed that the paint contained 2 
pounds of whiting and aniline color to 1 pound of red lead. This 
paint was completely disintegrated and ruined after only one year's 

exposure. 

A mixture of 70 per cent of barjrtes and 10 per cent each of car- 



MIXED PAINTS. 313 

bon black, zinc oxide, and amorphous graphite, well ground together 
in linseed-oil containing a small amount of Japan drier, will outwear 
red lead. If the coating is applied to a rusty surface, the scales of 
rust will break through red lead sooner than through the above 
mixture. 

Mixtures of zinc-white and white lead, both true pigments, are 
thought by. some engineers to be a more durable coating than either 
alone, and for some external exposures are said to be improved by 
the addition of 10 to 15 per cent of silica or barytes. 

The barytes and silica in these cases, also when added to flake 
graphite, save oil and give weight and bulk, as well as a frictional 
element to hold the graphite in place during the setting of the 
paint. 

In all these instances the paint appears to be better for the pres- 
ence of the various substances; but could the same group of mate- 
rials be combined into a single pigment, it would prove superior to 
any mechanically arranged article. 

The few cases where the mixture of true pigments with each other 
or with an inert substance has proved to be beneficial are not 
numerous enough to afford any foundation for a mass of incongru- 
ous substances called "mixed paints'* that flood the market, and 
which the reputable paint-manufacturer is almost powerless to stem. 
(See Chapters V-XXX.) 

No reliable paint can be made without skilled labor at almost 
every stage of its manufacture, even with the aid of the best me- 
chanical devices to reduce the labor account. Generally the labor 
and power account is two and a half to three cents per pound of 
paint, or twenty-five to thirty cents per gallon. Any paint worth 
appljring to a ferric or any other structure of importiince cannot 
be bought for forty or fifty cents a gallon. Such paint will not pro- 
tect a ferric body, and if applied will prove (because of frequent 
renewals) more expensive than one costing three times as much. 

Iron oxide is the cheapest straight pigment in use. But this 
cannot be properly ground in a reKable oil, barreled, and delivered 
for less than seventy cents per gallon, unless the manufacturer is 
losing money, or using an adulterated oil or a very unreliable iron- 
oxide pigment. 

Mixed paints containing resin, resin-oil, coal-tar in any form, 
calcium chloride or sulphate, and iron oxide act as carriers of 
oxygen and promote corrosion; hence they are unreliable coatings. 



314 MIXED PAINTS. 

The quality of the linseed-oil used in many of the mixed paints is 
against them. Oil made from unripe, smoky, condemned, or " no- 
grade '^ seeds, that often contain almost as much non-dr\4ng oil as 
the dr}'ing element, is too frequently used, and benzine is used for 
the drier instead of japan or turpentine. When linseed-oil is heated 
to 120° or 130° Fahr., and benzine is slowly added and well stirred, 
the mixture will not separate on cooling, but remain fixed until the 
oil is spread and dried by the evaporation of the volatile. The odor 
of the benzine in the paint in this case is almost suppressed; it is 
only markedly noticeable when the oil is again heated to near the 
above degree. 

Hot mineral oil added to linseed-oil is hard to detect by the odor, 
but the character of the mineral oil as a non-dr}ung oil is not changed 
in the slightest degree by the heating. A gill of mineral oil added 
to a gallon of red-lead paint will delay the setting of the paint. The 
paint never dries hard, but only on the surface. It remains viscid 
beneath, and the coating is liable to peel at any time. 

There were sold in the United States in the year 1900 60,000,000 
gallons of mixed paints and pastes, the use of which is increasing 
about 10 per cent each year. More paint is used in the United 
States per capita than in any other country. The English mixed 
paints are adulterated quite as much as any paints produced in 
the United States. 

The highly extolled English "Torbay'* paint is made from a 90 
per cent iron-oxide pigment, and has no merit over any American 
brand of oxide paint containing the same percentage of iron oxide 
and an equal quality of American linseed-oil. 

"Ferredor," an English trade-mark mixed paint, is exploited 
as being manufactured from a "natural metallic steel-gray ^ 95 per 
cerU rustless peroxide of iron found, in a very fine state of division, 
as a crystalline 'peroxide, and surpasses the oxides of iron, and is 
superior to red lead. 'Ferredor' cannot absorb or impart oxygen, 
so that the oil in the paint is not destroyed, as is the case with the 
red oxides," etc. 

The most brazen advertiser of any grade of American mineral 
brown paint never equalled this, and no American has been bold 
enough to attach to his product so fearful a trade-mark as "Schuppen- 
panzerfarbe." 

Crosbie's paint (English) is honestly stated as being made from a 
90 per cent oxide of iron, unadulterated and extra-finely ground 



MIXED PAINTS, 315 

in the best linseed-oil. Five governments and one hundred and 
thirty corporations show their confidence in an honestly made paint 
by using this firm's product, though it is but a corrosive one. 

"Armour-Scale Paint'' (Panzerschuppen). A Swiss ferric paint 
made from a number of formulae, also one or more German paints 
bearing the same trade-mark, are simply iron-oxide paints made 
from 80 to 90 per cent iron ores that the manufacturers call "granular 
micaceous." A greasy, crystalline, scaly iron ore with gangue, etc. 
Graphite in some (unclassified) form is added and the vehicle is some- 
times a linseed-oil varnish. 

"Lender's Anti-Corrosive Paint," especially recommended as 
impervious to heat, cold, warm water, steam, volatile acids, alkalies, 
gaseous ammonia, hydrochloric acid gas, and sulphuretted hydrogen 
gas. The base of the pigment is called "a silicate of iron." It is 
simply an ordinary iron ore containing iron-oxide, 88.66 per cent; 
silica, 6.40 per cent; lime and magnesia, 3.10 per cent; alum and 
phosphoric acid, 0.55 per cent; undetermined substances and loss, 
4.39 per cent. It is sold in the form of a paste, being finely ground 
in a boiled oil or varnish, and when used is reduced with raw linseed- 
oil, and litharge added for a drier. A special point claimed for this 
wonderful oxide-of-iron paint is that "any mineral paint of the 
right color can be added to produce the desired tone." Won- 
derful product! The special advantages for this special paint are 
accompanied by a specially high price for it. 

Other examples of the foreign mixed paints and the art of adver- 
tising them could be cited. A mixed paint is not necessarily a good 
one because of its trade-mark and high price, nor is it a notoriously 
bad one because of its being an American firm's product. 

Some instances of the unreliable character of compound paints 
have been given in Chapters Y (White lead) and XXX (Paint 
tests). The following instance, from the cost and magnitude of the 
structure to which the paint was applied, the public interests at 
stake, and the result of the application, is of interest. It illustrates 
the unreliable character of compounded pigments applied over 
other basic pigment paints for the protection of ferric stmc- 

tures. 

The new suspension bridge over the East River has in Brooklyn 
a viaduct about 9000 feet long. This viaduct is of lattice trusses 
and columns of the type used for elevated railways. 

The steel for this structure was specified and supposed to have 



316 MIXED PAINTS. BROOKLYN BRIDGE VIADUCT. 

been cleaned by the sand-blast, but it mostly received only a coating 
of boiled oil before machining. After riveting up in sections pre- 
paratory to shipment it received a coating of red-lead paint, and 
after erection another coating of red lead and lampblack paint was 
applied. 

After a number of months' drying the first finishing coat of Jewett's 
white lead, 70 to 75 parts, and the New Jersey zinc oxide, 20 to 25 
parts, ground in raw linseed-oil, was applied. The engineer corps, 
from some previous experience with this mixture, were particular 
to see that the quality of all the materials was standard in all respects. 
The spreading of this and the following or fourth coat was in situ 
and under their eyes; hence they alone are to blame for any errors 
in this. After a number of months, the fifth or finishing coat, com- 
posed of the above white lead and zinc oxide mixture, also a small 
quantity of French ochre to make the coat cream-colored, was ap- 
plied. 

In a not particularly severe exposure, and in two years, the last 
three lead and zinc coatings were disintegrated and washed away 
in large areas over the entire structure, showing the foundation 
coats of red lead. A cloth wetted in water and wiped over the coat- 
ings washed them off as freely as though they were of whitewash. 
Wringing the cloth left the three coats of lead and zinc pigments in 
the water like so much chalk. If the percentage of zinc oxide in 
the coating had been forty, the paint would have failed by peeling 
in strips instead of chalking. The paint, in its materials, propor- 
tions, application, location, and exposure, is not far diflferent from 
that of the New York City elevated railways, which thus far has 
kept its place free from chalking, but has not prevented corrosion 
from attacking every foot of the structure beyond the possibility of 
correction. 

Had the same amount and quality of the paint been spread on 
a thousand inland structures, the failure of the coatings would not 
have been so marked as in the above case, where the agency of acres 
of decayed paint tells the tale; but it would have occurred just the 
same, though the loss would have been so widely distributed as to 
call no special attention to it. 

The reliable qualities of a paint composed of a number of 
pigments, or a compound paint, where the substances of the com- 
pound are united as an individual whole by chemical affiliation 
in the process of manufacture, have been mentioned a number of 



MIXED PAINTS 317 

tim«s in this work. The annexed figure (46) iUustrates the dura- 
ble natiu« of that class of pig- 
ments. 

It is the photograph of a 
four^inch-diameter wrought-iron 
pipe. The upper end, A, was en- 
closed unpdnted in a cast-iron 
ring. The lower end, B, is the 
adjoining part of the pipe that 
had been painted with two coats 
of sublimed lead and zinc. The 
pipe and its connections were 
buried in the earth for nine years. 
When taken up the end A had 
corroded and lost over ^^ inch 
in diameter. The part B was 
uncorroded and nearly all of 
the paint on the whole length 
of pipe was unchanged and in 
place. Where the coating was 

scaled off in the process of re- y,^ 48.— Four-inch wiought-iron pipe, 
movingthe pipe from the trench, 

the iron was as clean and uncorroded as when laid. (See Chapter 
V, Sublimed Lead.) 

Sir Benjamin Baker has concisely remarked that "it is the 
deviation from the average which really is so important in the design 
of en^eering works." 

This is equally applicable to the design of a paint. The majority 
of paint^manufacturers appear to harbor the idea of the universal- 
ity of their product. They give little or no attention to location, 
exposure, or the many influences to which it may be subjected during 
ita life on a ferric structure. 

A paint that is Triable m open inland locations often fails on 
the aeaeoast, or in manufacturing towns, or on industrial establish- 
m^ts. The paint that is reliable in the latter place may fail quickly 
on another not far distant location because of disregarding one or 
more of the above factors. 

It is important that both the engineer and the paint-eompounder 
remember that there are a number of affinities in a paint coating. 
These are where the vehicle and the pigments are capable of mixing, 



318 ENAMEL PAINTS. 

not to form a new chemical compound, but to influence the action of 
one of the group; for instance, where red lead or red lead and 
lampblack are added to influence the drying of the paint. Another 
instance is where a substance combines in definite proportions, such 
as the carbonate of lime added to an iron oxide to neutralize the 
sulphm* in the pigment, or the acid in the vehicle. 

It is also to be kept in mind that every metal is electropositive 
to its own oxide; the latter induces corrosion in the former when- 
ever the two are brought into contact, as in a paint coating. The 
vehicle only insulates or protects either substance but indifferently, 
particularly where the covered metal is the base of the oxide in the 
paint. The water and acids in the oil (the latter not infrequently 
rancid) have great influence in the decay of the coating and corrosion 
of the covered metal. 

The production of enamel paints has become a trade of great 
importance. They differ from ordinary oil-vehicle paints inasmuch 
as they dry with a high lustre or gloss. They were first called "var- 
nish paints," for the reason that the vehicle was a varnish, instead 
of linseed-oil. They are used principally for coating walls on the 
inside of buildings that require to be washed with water, as in hos- 
pitals, courtyards, etc., the varnish vehicle being less easily injured 
by the water than an oil vehicle. 

The use of a varnish vehicle for ferric paints has been tried in a 
number of instances in late years with varying degrees of success. 
The principal difficulty in their use is the uncertain character of the 
varnish vehicle, which requires a greater knowledge of the nature 
of the fossil resins and how to compound them than the average 
paint-manufacturer has at his command. 

There are three classes of enamels: * 

First. The slow-drying enamels, that require from twelve to 
fifteen hours to dry under normal conditions. They give a fine, 
lustrous, level coating, and if carefully applied should show no brush- 
marks. They are essentially an oil and fossil-resin varnish, witii 
which the necessary color pigments are ground. 

Second. Quick-drying enamels, that dry in from twenty to forty 
minutes according to their composition. They are essentially spirit 
varnishes, colored with pigments. They leave a lustreless or flat 

♦"Enamel Paints and how to use them." By Geo. W. Hurst, F.C.& 
Western Paint Magazine, Feb. 1900, pp. 46, 4Z 



ENAMEL PAINTS. 319 

surface, are apt to show brush-marks, and are not as diu^able as the 
slow-drymg enamels. 

Third. Baking enamels. They are varnish compounds that when 
heated flow into a uniform and lustrous coating. The pigments 
added to them give them body and color. Sewing-machine and 
bicycle frames, hardware, etc., are examples of these coatings. In 
larger ferric bodies they are represented by the baked japans ap- 
plied to water-pipes. (See Chapter XI.) 

The composition of the varnish base of all of these enamels is 
the essential part, and in most cases better results can be attained 
if the varnish is obtained from a reliable varnish-manufacturer than 
where the painter makes it himself. Even the best of varnish- 
makers fail sometimes to produce a reliable varnish owing to many 
causes, principally from the use of common resin and other poor- 
quahty gums, the effect of which is the " crazing " of the coating. 

The general nature of the varnish base is indicated in the follow- 
ing recipes that have given good results: 

Sixty poimds of good, white Sierra Leone copal, mixed with 10 
gallons of the best quality of hot boiled linseed-oil. When well 
cooked add 16 gallons of turpentine and ^ pound of linoleate of 
manganese for a drier. This varnish, suitable for all colors from 
black to white, should be ground in as for oil paints. 

A cheaper grade of varnish can be made from 35 pounds of Kauri 
gum and 15 pounds of Sierra Leone copal, mixed with 7 gallons of 
hot boiled linseed-oil and 1^ gallons of turpentine. When nearly 
cold, thin down by adding 10 gallons of turpentine. This will be 
too dark for white enamels, but answers for all other colors. 

The white enamels are usually made by adding zinc-white or 
lithopone to the varnish. They work well. About 6 pounds of the 
white is required to a gallon of the varnish. Whiting and pipe-clay 
are detrimental; they make the coating gray, are deficient in body, 
and are liable to cause "peeling." 

Black enamels require 4 pounds of lampblack to about 6 gal- 
lons of either grade of the above varnishes. 

A quick-drying varnish base is made from Sandarac gum, 10 
pounds; soft Manila copal, 5 pounds; gum benzoin, 1 pound; 
methylated spirit, 8 gallons; or 4 gallons each of methylated spirit 
and wood-naphtha. This varnish is suitable for all colors from white 
to black. A cheaper varnish, suitable for dark colors and black, is: 
Sandarac, 12 pounds; shellac, 10 pounds; benzoin, 2 pounds; methyl- 



320 ENAMEL PAINTS. 

ated spirit, 12 gallons. These varnishes dry quickly. A slower- 
drying varnish is made from gum dammar, 14 pounds, and 3 gallons 
of turpentine. 

For a white enamel from either of these varnishes grind in 10 
pounds of zinc-white or lithopone to 1 gallon of varnish. 

For a black enamel grind in 5 pounds of zinc-white, 2 pounds of 
carbon black, and 3 ounces of brilliant ebony spirit-black to 1 gallon 
of varnish. 

The baking enamels are not essentially diflferent from the first 
class of varnishes herein mentioned, other than that they contain 
more gum or resin of some quality. 

A few special recipes are the following: Best grade of refined 
asphaltum, 70 pounds, mixed with 9 gallons of hot linseed-oil and 5 
gallons of gold size. Boil until ropy, then add 9 gallons of turpen- 
tine. This is for a black coating. 

The colored enamels require a better grade of varnish base, which 
is made thus: A good quality of Kauri and copal gums, each 20 
pounds; animi, 6 pounds; melt together and mix with 14 gallons of 
hot linseed-oil; boil until stringy, thin with 18 gallons of turpentine, 
and use with any pigments to get the desired color. 

A quick-drying black enamel is composed of: 

D. C. shellac 60 pounds 

Gum sandarac 20 ** 

Gum benzoin 2 " 

Lampblack 5 " 

Castor-oil 0.75 " 

Spirit aniline black 1 . 50 " 

Methylat-ed spirit. 26 gallons 

Wood-spirit. 2 " 

Total 89.25 pounds 27 gallons 

The pigments assembled with all of the above varnishes should 
be of the best quality, particularly the lampblack. Pulverized 
bituminous coal and nearly all of the so-called "carbon blacks" 
prove detrimental to the quality and color of the enamel. No ben- 
zine or turpentine can be added to any of the above varnishes after 
they have left the cooking-kettle without affecting the gloss of the 

enamel. 

The above enamel compounds represent only a few that are put 
upon the market under many special names. As a rule, the pro- 



ENAMEL PAINTS. 321 

tective value of all enamel paints is quite as variable as the ordinary 
oil- vehicle paints. A varnish vehicle does not always secure a reliable 
ferric coating, unless more care is exercised in preparing the surface 
to receive it, or in its application, than the average painter generally 
gives to these matters. The use of resin or resin-oils in enamel paints 
is as disastrous as when used in an oil paint. 



CHAPTER XXXIII. 

CORROSION OP IRON AND STEEL. 

The difference in the rate of corrosion between iron and steel 
as given by different authorities varies greatly. The reason for 
this is plain: the conditions of each reported rate of corrosion, whether 
the result of laboratory or other tests, such as exposure to weather 
or other corroding influences, not being similar in all respects, are 
therefore comparable only in a general way. 

Pieces of iron and steel, both suitable for boiler-tubes, were made 
clean and bright, then placed in sandy loam with which had been 
thoroughly mixed some sodium carbonate, sodium nitrate, ammo- 
nium and magnesium chlorides. The earth so prepared was kept 
moist. At the end of twenty-three days the plates were taken out, 
cleaned and weighed. The following was the result: 

Iron had ost by corrosion 0.84 per cent. 

Steel " " '' '' 0.72 " " 

The pieces were replaced in the earth and left for twenty-eight 
days longer, or sixty-one days in all. The result was: 

Iron total loss by corrosion 2.06 per cent. 

Steel " " " " 1.79 " " 

This is a rate of corrosion that would probably have caused the 
disappearance of the plates inside of eight years. 

Experiments conducted by the Admiralty, Board of Trade, and 
Lloyds prove that steel corrodes much more rapidly than iron 
when exposed to the action of salt water; also that the commoner 
brands of iron corrode less rapidly than the better brands when 
exposed to the same influences. With steel and iron both unpro- 
tected and exposed to the same action of the weather and sea-water 
corrosion advanced at the rate of one inch in depth in 82 years 
for the steel and 190 years for the iron. When always immersed in 
sea-water the periods are one inch in 130 years for the steel and 310 
years for the iron. When always immersed in fresh water the 

periods became 600 yeais for the steel and 700 years for the iron. 

322 



CORROSION OF IRON AND STEEL. 323 

Mr. B. H. Thwait, A.M.I.C.E., reports that a bar of wrought iron 
tinprotected, exposed to the action of the atmosphere in a manu- 
facturing town, demonstrated that a bar of common iron one inch 
by four inches would be entirely corroded away in a little over 100 
years. 

Mr. G. Rennie's experiments in 1836 were with cubes of wrought 
iron, cast iron, and bronze for lighthouse purposes. The cubes in 
separate vessels were immersed for seventy hours in saline solutions 
considerably stronger than sea-water. The cast iron lost yg*^ of its 
weight; the wrought iron ^^y^r, being in the proportions of two to 
one in favor of cast iron. The bronze lost -uftiru o^ ^^ weight, a result 
in favor of bronze over cast iron of three to one. The cast- and 
wrought-iron cubes were then placed in a strong solution of one part 
of muriatic acid in twenty-five parts of Thames water, and exposed 
for twenty-one hours. The cast-iron cube lost ^ of its weight, the 
wrought-iron only y^, being eight to one in favor of the wrought 
iron. 

In these experiments with the same samples of each metal the 
results were directly contrary. The crystalline nature of the cast 
iron evidently favored the disruption of the crystals from their bond 
or loose association together; they were cast off when partially cor- 
roded, and corroded by themselves while the acid had fresh sur- 
faces to act upon in all directions. 

The experiments of Mr. Robert Mallett, M.I.C.E., on wrought 
iron and cast iron sunk in the sea, showed that from ^ to j*^ inch 
in castings one inch thick, and about ^ inch of wrought iron, will 
be destroyed in a century in clear sea-water. This is equal to fifteen 
to one in favor of cast-iron. Other experiments by Mr. Mallett 
showed that cast iron unprotected and exposed freely to atmospheric 
action was corroded nearly as rapidly as by the action of clear sea- 
water. 

Mr. Kenniple, Central India Railway Company, reports that 
"the greatest corrosion of cast-iron piles existed close to the low- 
water mark, and did not extend to any considerable distance from 
that point." This condition he also found to exist in the wrought- 
iron bolts and braces. After an exposure of twenty-five years the 
piles were found to be in very good condition, and corrosion had only 
occurred in places accessible for renewals. A thin coating of mud, 
marine growth, and barnacles upon the immersed surfaces of the 
ironwork that protected them from contact with fresh supplies of 



324 



CORROSION OF CAST-IRON CUBES. 



water, or from the water in motion, had a tendency to retard cor- 
rosion, but when they were removed corrosive action increased at 
once. He concludes that after a life of from thirty to fifty years 
cast-iron structures exposed to sea-water can only be regarded as of 
a temporary character, especially those of light cast-iron pile design. 

Dr. Grace Calvert, F.R.S.,* experimented on a number of gray 
cast-iron cubes made of Staffordshire cold-blast iron immersed in 
acidulated water. The specimens were 0.39 inch cube, specific grav- 
ity 7.858, weight of cube 237 grains. The cubes were placed sepa- 
rately in bottles holding about 30 cubic inches of greatly diluted 
sulphuric acid. Similar cubes were placed in bottles containing 
dilute hydrochloric, acetic, and phosphoric acids. The action of the 
acids on the iron was slow; but at the end of three months, although 
the appearance of the cubes had not changed, some of them, es- 
pecially those immersed in the solution of acetic acid, had softened 
so that a knife-blade could penetrate them 0.11 to 0.16 of an inch. 

The solutions of acids mentioned were replaced by a fresh one in 
each bottle every two months for two years. Changes were then 
found to have taken place in all of the cubes, the acetic acid show- 
ing the greatest decomposing effect, then the hydrochloric, sulphuric, 
and phosphoric acids; the latter had the least effect upon the cubes. 

The action of the acids upon the iron had changed its nature, 
without any alteration of its bulk or in the appearance of the sur- 
face of the cubes. The weight of one cube after two years' immer- 
sion was 54 grains against its original weight of 237 grains. Its 
specific gravity was 2.751 instead of 7.858. The change in the physical 
charact^er of the iron is indicated by the following analysis of the 
gray cast-iron from which the cubics were made, and a set of cubes 
after a two years' immersion in the acetic acid solution: 





Before Immersion. 


After Immersion. 


Iron 


. . 95.413 per cent. 


79.960 percent. 
11.070 " 

2.590 " 

6.070 " 

0.096 " 

0.059 " 

0.155 " 


Carbon 


2.900 " 
0.790 " 
0.478 " 
0.179 " 
. 132 " 
0.108 " 


Nitrogen 

Silicon 


Sulphur. 

Phosphorus 

Loss. 


100.000 


100.000 



* Minutes of Proceedinf^rs of the Institute of Civil Engineers. 



CORROSIVE INFLUENCES IN THE ATMOSPHERE. 325 

Dr. Angus Smith found that, taking the inland countty parts 
of England as a basis for the acidity of rain-water and the impurity 
of the atmosphere, and rating them as 0, in Glasgow they were 83, 
and in London 28. 

The comparative amounts of ammonia and other impurities in 
the air and in rain-water were: Valentia 1, Glasgow 50, Liverpool 
30, Manchester 36. For the amoimt of hydrochloric acid present in 
the same elements Blackpool was 100, London 320, the Under- 
ground Railway in London 974. Anhydrous sulphuric acid at 
Blackpool 100, London 352, Underground Railway 1554. Ammo- 
nia and albimiinoid ammonia at Innellan, on Firth of Clyde, 100, 
London 108 and 117, Glasgow 150 and 221, the Underground Rail- 
way 138 and 271. 

Drs. Clowes and Andrews' examination of the air in the cars of the 
Central London Railway showed a maximum amount of carbon dioxide 
of 14.7 volumes and a minimum amount of 9.6 volumes in 10,000 vol- 
umes of air. In a railway-station elevator 15.2 volumes was found. 

Many points in the Paris underground railway system have 33 
volumes of carbon dioxide in addition to other deleterious gases and 
2 per cent of aqueous vapor. Dr. Clowes states that not more 
than 8 volumes of carbon dioxide in 10,000 volumes of air should ever 
be present at any point in a railway building. 

Dr. Smith also found a variety of solid substances in the air, such 
as common salt, sulphur, nitrate of ammonia, lime salts, iron; ako 
the phosphates, iodides, and other organic matters given off by ani- 
mals, vegetables, etc. The percentage of oxygen in the open air 
varied from 21.0 to 20.40, while the carbonic gas varied from its 
normal amount of three parts in ten thousand to over thirteen parts 
in the London Underground Railway, and some parts of the Swiss 
tunnel contained over 17 per cent. 

Dr. Huxley's Physiography gives his examination of the amount 
of carbonic acid gas in 10,000 parts of the atmosphere at a number 
of points a& examined by him, viz. : 

On the River Thames at London, mean 3 . 43 

In the streets of London " 3 .80 

Top of Ben Nevis, Scotland, 4436 feet high, mean. 3 . 27 

A waid in St. Thomas' Hospital, London 4 .00 

Haymarket Theatre, London, Dress Circle at 11.30 p.m. when 

lighted bv gas 7 . 57 

Underground Railway, London 14 . 52 

Average of 339 English mines 78 . 50 

Highest amoimt in a Cornish mine 250 .00 

See Chapter XXXVI for corrosive elements in anow-water. 



326 CORROSION OF CAST-IRON PIERS. 

Dr. W. G. Black* gives the results of his examinations for dust 
and soot in the air, in the central district of Edinburgh during the 
year 1902. The fall of dust and soot in an open dish of 75 square 
inches area amounted to 2 ounces, equal to 3.8 ounces per square 
foot, or 23.5 pounds for every 100 square feet. 

The above amounts of corrosive elements are not arbitrarily con- 
stant, but they indicate the corrosive influences that may be encoun- 
tered in almost every situation of engineering work. 

Mr. Beardmore, C.E., instances a case of a sea-lock, in which 
soft water was "locked" down into the sea-water level. At the end 
of thirty-five years' service all of the cast- and \\TOUght-iron attach- 
ments to the wooden lock-gates, also the spikes in the platforms and 
gate-sills, had completely corroded away, though the timber parts 
of the structure were perfectly sound. 

Ferric metal exposed to the action of salt or fresh water which 
is not changed corrodes far less rapidly than where the water is 
changed more or less frequently. 

Dr. Lyon of India, a high chemical authority, reports that 
some cast-iron piles, after four and a half years' exposure to the 
action of pure sea-water having a specific gra-vity of 1.028 and that 
contained 3000 grains of solid matter per gallon, of which 1605 con- 
sisted of the chlorides of sodium, magnesium, etc., had undergone 
a change to a depth of -j^^ inch from the surface of the metal. 

The Milton-on-Thames pier was erected in 1844 on cast-iron 
columns 3 feet in diameter and 1^ inches thick. In the Gravesend 
town pier and the Maplin Sands lighthouse the cast-iron columns 
and other cast-iron members exposed to the sea-wat^r were originally 
IJ and li inches thick. In all of these structures, at the end of forty- 
five years, only f inch of the metal remained unaffected; the rest 
had changed to the semblance of plumbago. None of the members 
of these structures indicated any change in the metal. 

Mr. Thomas Rhodes, C.E., reports that "in the locks of the 
Caledonian Canal the cast-iron sluice-gates were exposed to sea- 
water. All of the parts were coated with a heavy Swedish tar, 
except on the working faces of the gates. These faces were faced 
and ground together. Four years after the immersion of the gates, 
upon an inspection of their condition, on all of the parts coated with 
the tar no corrosion was apparent, while the machined and ground 



* Royal Meterological Society Journal, XXII, 1903, p. 134. 



CORROSION OF CAST-IRON ARCHES. 327 

working faces were softened and changed to plumbago to the depth 
of I of an inch, and had to be renewed." 

The experience of American engineers appears to be equally 
conclusive of the treacherous character of cast iron exposed to sea- 
air or sea-water. 

Mr, John D. Van Buren, in a paper read before the American 
Society of Civil Engineers, stated that "bolts and other wrought-iron 
parts are badly corroded in less than twenty-five years when submerged 
in sea-water. Certain kinds of cast iron could perhaps be made to last 
fifty years, which would be a generous allowance, and probably 
greatly exceeds the average life of cast iron exposed to sea-water." 

Sir Benjamin Baker, in his paper "The Metropolitan and Metro- 
poUtan District Railways," * says: "In timnel constructions when the 
roof members and bottom flange of the girders, tie-rods, anchors, etc., 
are much exposed to corrosive influences, wrought-iron members 
were used and thought to be more trustworthy than cast iron, but 
were found to be exposed to a greater risk from hidden oxidation. 
Experience has shown the trouble and cost of maintaining ironwork 
exposed to atmospheric corrosion in an underground railway. It is so 
great that it would justify a considerable increase in the first cost by 
substituting brickwork and deep cuttings for ironwork and shallower 
construction. Where the depth was sufficient for an arch, brick 
covered ways were to be preferred to iron-girder constructions, on 
account of the smaller cost, increased durability, and safety." 

Sir John Fowler confirmed Sir Benjamin Baker's views in regard 
to the substitution of brick work and masonry for ironwork in all cases 
where possible, even at a material increase in the cost of the work. 
He thought the question was not confined to the relative rate of corro- 
sion of cast iron, wrought iron, or steel, but to the great risk arising 
from hidden oxidation on important members of walled-in ironwork. 

These views are corroborated in the experience of American 
railways. The Pennsylvania Railway has already replaced a num- 
ber of its iron bridges with masonry arches, and wherever possible 
will continue the substitution. 

Car-wheels made from the best of gray cast iron when inunersed 
either constantly or alternately wet and dry, in fresh or salt water in all 
stages of impurity are found to corrode faster in the body part of the 
wheel than in the tread or chill, the iron being identically the same 



♦ Minutes of ProceedingB of the Institute of Civil Engineers, Vol. LXXXI. 



328 CORROSION OF GRAY CAST-IRON PIPES. 

in both parts of the wheel and under the same corrosive influences. 
The difference must be attributed to the effect of the chill, which 
has changed, the molecular formation of that part of the wheel, 
making it more dense and of a needle-like or filamentous formation, 
that is not so readily attacked by corrosion as the crystalline part. 
This effect in car-wheels is the more noticeable as the body part of the 
wheel is covered with a skin of the sihcate protoxide of iron, produced 
by the molten metal fusing the sand in the mold. There is also a further 
protection afforded by a film of the black magnetic oxide of iron being 
formed at the same time by the oxidation of the hct metal in cooling. 
Cast-iron pipe or piles would be more durable were they cast from 
chills the same as the tread part of a car-wheel. Instead, however, 
of this method, which is attended with some difficulties not present 
in the casting of car-wheels, it has been proposed to increase the 
depth of the silicate coating on such bodies by a process used by the 
ate Mr. E. F. C. Davis, late President of the American Society of Me- 
chanical Engineers, in the protection of mining pumps, chambers, 
and pipes subject to the action of mine-water. The corrosive action 
of this fluid is very great and increases by motion and pressure. The 
process consists of coating the cores and other parts of the mould 
with a thin paste appUed by a brush, which increases the thickness 
of the silicate coat to more than double its natural thickness, and 
can be made still thicker by repeated apphcations of the paste before 
casting. The composition of the paste is, viz. : 

112 parts silica. 

44 " calcined soda carbonate. 
24 " " carbonate precipitate. 

4 " boracic acid. 

Mix and thoroughly pulverize them. Coat the core or mould with 
plumbago facing, and apply the enamel as a powder or paste to the 
thickness required, in one or more applications, and cast as usual. 
A single coating with this compound more than doubles the Ufe of 
the metal. 

Fresh or sea water impregnated wdth decomposing organic mat- 
ter, or acids and chlorides discharged from bleacheries, paper-mills, 
etc., hastens corrosion, which is increased by motion, pressure, and 
high temperature. 

The Merrimac River, at Lowell, Mass., is affected to such an ex- 
tent by the discharge waters from manufactories, that during the 
ordinary flow of the river in summer, it will change Htmus paper. 



CORROSION OF GRAY CAST-IRON AND CANNON, 329 

Instances are on record where finn gray cast-iron water-pipes of 
krge diameter and 1 inch in thickness, conveying sea-water, have 
inside of five years changed to a phraibago-like substance, that had 
hardly any strength and could be cut easily. No change in the appear- 
itnce of the metal was indicated ; the change in the metal was from the 
inside, or where in contact with the water. The pipes had the usual 
tough skin, and were nominally protected by a foundry dip coating. 

Mr. F. A. Boyer, M.E.,* reports that a cast-iron water-pipe 12 inches 
in diameter, 1200 feet long, used to circulate sea-water for conden- 
sation purposes, laid mostly in the open air, was changed in two and 
one-half years to a plumbago-like substance that could be cut easily, 
but to the eye presented no indications of the change in the metal. 

Gray cast-iron water-pipes f laid at Atlantic CSty, N. J., for nine- 
teen years, in black swamp-mud containing decomposed vegetable 
and saline substances highly acidulous and corrosive, were found 
to be generally corroded externally. The iron was softened about 
i inch in depth, and in many places it had extended through 
the pipe. The corroded metal appeared to have been replaced by 
a clay-like substance of light weight, containing 17 per cent of silica, 
particles of the cast iron being embedded in it. 

A member of the American Water-works Association reported 
in the 1902 meeting the decay of two cast-iron suction-pipes, where 
the metal was changed to a graphitic substance, easily cut. The suc- 
tion water was originally soft groimd-water, but later was artesian 
well-water, containing a small amount of sulphuretted hydrogen. 
Wherever brass and iron came in contact with the water-mains, 
eventually the iron became soft enough to cut easily. 

Mr. Trautwine, an eminent American civil engineer, recommends 
a white close-grained cast iron or chilled iron for piles and wharfs 
exposed to sea-water or sea air. 

Mr. Turner, Assoc. R.S.M., recommneds a gray cast-iron for 
docks and piles exposed in English sea locations. 

The above results of the use of different brands of cast iron for 
sea-air and sea-water exposures in different locations are probably 
due to the difference in sea-water in different parts of the world. 
Sea-water varies greatly in corrosive and fouling properties even if 
not contaminated as mentioned before. 

* TransactioDs American Society of Mechanical Engineers, Vol XVI, 1894. 
Paper No. 626, p. 416. 

t Prof. W. P. Mason, Troy Polytechnic Institute, Troy, N. Y. 



830 RATE OF CORROSION BETWEEN IRON AND STEEL. 

Cannons and other tough cast-iron articles, 6 or more inches 
in thickness of metal, sunk in the sea in many parts of the world, 
change in about one himdred years to a soft carburet of iron or a 
plumbago-like substance, without any diminution in size or any 
exfoUation in scales or flakes, as in atmospheric oxidation. They 
become too hot to handle, when first raised and exposed to the air, 
from the absorption of oxygen. 

Mr. Francis T. Bowles, Naval Constructor United States Navy, 
states: "That the corrosion of iron and steel, from the observations 
made at the different Government navy-yards, with the different kinds 
of iron and steel used in naval vessels exposed to sea-water in various 
parts of the world under a great number of conditions of tempera- 
ture, brackish, sewage, and dock water, was: That unpainted iron 
and steel plates will corrode in one hundred years on each exposed 
surface .30 to .50 inch of metal; in ordinary fresh water, .02 to 
.03 inch, and in the atmosphere, .25 to .30 inch." But he makes no 
distinction in the rate of corrosion between iron and soft steel. 

Mr. L. M. Hastings, City Engineer, Cambridge, Mass., reporting 
the results of his experiments to ascertain the difference in corro- 
sion between wrought iron, soft steel, and cast iron, all uncoated, 
exposed to a running mixture of pond and brook water, fairly soft 
and comparatively free from any acid or saline matters, states: "That 
after an exposure of one year the uncoated wrought- iron was badly 
tuberculated and rusted, the soft steel was similarly affected, but 
not to the same degree; the cast iron was also affected. The per- 
centages of increase in weight due to corrosion were as follows: The 
wrought iron, 1.5 to 3.1 per cent; the steel, 1.1 to 2.1 per cent, and 
the cast iron, 1.0 to 0.96 per cent. A similar set of plates buried 
in sand, also in clay soil, showed the same relative difference in cor- 
rosion of the metals." 

Dr. Robert H. Thurston* reports the result of his investigations 
of the different rates of corrosion between iron and steel. Briefly, 
they are: "Cast iron in dilute solutions of acids is rapidly acted 
upon, the metal retaining its general appearance unchanged. 
The condensation waters from engines are strongly corrosive. Hard 
iron, rich in combined carbon, rusts slowly. Graphitic iron, also 
different qualities of iron in contact, increases the rate of oxida- 
tion presimiably by forming local voltaic couples. Hard steel 

* "Materials of Engineering," Sec. 192, Vol II, pp. 328 et seq. 



RATE OF CORROSION BETWEEN IRON AND STEEL. 331 

rusts less rapidly than soft steel. Bilge-water corrodes iron and 
steel rapidly. Iron ships carefully painted have been found to cor- 
rode at a rate not far from ^ inch in twenty-five years." 

Thwaite* gives a formula and table of constants for the rate of 
corrosion between different metals under different elemental expo* 
sures, all for unprotected metal, viz.: 





Water. 




Material. 


Sea. 


River. 


Impure Air. 




Foul. 


Clear. 


Foul. 


Qear Water, 
or in Air. 




Cast iron 


0.0656 
.1956 
,1944 
.23 
.09 


0.0636 
.1255 
.0970 
.0880 
.0359 


0.0381 
.1440 
.1133 
.0728 
.0371 


0.0113 
.0123 
.0125 
.0109 
.0048 


0.0476 


Wrought iron 

SteeL 


.1254 
.1252 


Cast iron, no skin. . . 
Galvanized iron. . . . 


.0854 
.0199 



Average for sea-water: Cast iron, in contact with brass, copper, or gun 
bronzes, 0.19 to 0.35; wrought iron, in contact with the same, 0.3 to 0.45. 

No analysis of the constituents of the several metals is given, and 
the terms hard and soft metal are very variable conditions. How- 
ever, all of the' above experimenters give hard metal, whether of 
cast iron, wrought iron, or steel, as being less corrosive than soft 
metal. 

Mr. Thomas Andrews, F.R.S., experimented at the Wortley Iron 
Works on wrought-iron and steel plates containing varying amounts 
of carbon. The plates were immersed in sea-water that was changed 
monthly. It was found that the lower the percentage of combined 
carbon in the metal, the lower was the corrosion. The best wrought 
iron corroded less than any of the steels at any stage of their expo- 
sure during the one hundred and ten weeks of the test. Wrought 
iron that contained double the usual amount of phosphorus and 
manganese corroded more than the iron free from these substances, 
but the corrosion in them was less for the whole period of the test 
than in any of the steels, with the single exception of a very soft 
Bessemer. 



Engineering News^ Nov. 3, 1898. 



332 CORROSION OF WROUGHT-IRON. 

In any of the steels, manganese in excess tended to produce an 
increased corrosion, evidently from its unequal distribution and 
the galvanic action on the adjacent metal. 

On the whole, from the many reported cases of corrosion in all 
parts of the world, that include many qualities of cast iron and 
wrought iron that have had approximate exposures for thirty or 
more years, it appears that hard close-grained and chilled iron are 
less liable to corrode than any brands of softer metal. Bessemer, 
open-hearth steels, also steel castings used for structural work, have 
not yet had time enough to afford many comparisons with each other, 
or with cast iron and wrought iron, to prove which metal is the most 
affected by corrosion. 

It is left to the future to develop some alloy of iron or steel that 
will retard if not prevent corrosion, while not materially reducing 
their strength or other qualities. 

The composition of wrought iron and the processes it is subjected 
to between the bloom and finished article have a great effect to 
determine its rate of corrosion. Iron containing sulphur is red short, 
that containing phosphorus is cold short. Both differ in corrosi- 
bility; the cold short is the one less affected, being harder and more 
crystalline in composition, while the red short has the sulphur ele- 
ment to aid corrosion. Neutral iron made from both of the above 
brands has a different rate of corrosion than either. 

The same quality of iron worked in the rolls, in the one case both 
lengthwise and crosswise, to produce sheet or plate iron, differs in 
corrosibility from that worked principally in one direction, as in 
the case of beams, angles, and other structural shapes. All of the 
latter forms tend to disintegrate by corrosion into strips, needle 
or fibrous form, owing to the granular character of the iron being 
changed by the rolls into parallel fibres, that are not int^erlocked 
as the cross-rolling arranges them. The corrosion aided by the 
cinder follows the grain of the metal. The cinder is acid and porous, 
and only in mechanical bond with the iron by reason of the action 
of the rolls. 

Fig. 47 shows the effect of laminated corrosion of a steel-plate 
girder on the Washington Street railway bridge in Boston. Wrought- 
iron pipe used for water, steam, and gas service is an example of 
this make of iron. It corrodes more rapidly than the same quality 
of iron in bars. The iron is not so condensed in the process of roll- 
ing a tube as in rolling a bar. The tube skelps receive their principal 



CORROSION OF WROVGHT-IBON. 333 

rolling lengthwise, while bar iron gets some edge-rolhng in passing 
through the rolls. Enough, in fact, 
to show a marked difference in the 
corrosion of the two products, when 
from the same metal. Boiler-tube 
skeipe, being made from a better 
quality of iron, or having been re- 
fined by a further working of the 
same grade of metal, by having had 
some cross-rolling before being made 
into skelps, are less affected by cor- 
rosion than ordinary wrought-iron 
pipes. The arger sizes of these pipes 
being made from long rolled skelps 
and lap-welded , instead of only welded , 

as in the case of small pipes, show in Fio. 47. — Lnminated coimsion 
the wel* a l<« corrcibility th.„ u. iX'^'"l,^" ZX 
the body part of the pipe. This is bridge in Boetan. 
due to the additional condensation of 
the fibres at the welds; also there is lees cinder in the lap-weld. 

Cold-rolled shafting and rods are rendered more dense by the 
roiling process, and they are lees affected by corro^on .than the bats 
from which they are cold-rolled. The process also increases their 
tensile strength. Cold-drawn wire also presents the same features. 
A wire nail corrodes less than a cut nail, so does a hammered or so- 
called wrought nail. 

Polished iron and steel tools, sword-blades, razors, etc., resist 
corrosion better than the same articles ground, but not polished; 
the improved resistance being due to the surface not having so many 
small cavities to hold any moisture reaching it, because of the con- 
densed and repellent nature of the surface due to the polishing. 

Burnished surfaces resist corrosion and are more repellent of 
moisture than polished ones; but, if corrosion is once established 
on them as a spot, it appears to concentrate an energy to produce 
a deep corrosion, that is difficult to eradicate. 

Rivets have a different rate of corrosion between their heads, 
or points, than the body of the rivet. Corrosion of a riveted joint 
generally concentrates its action more immediatdy around the rivet 
heads and points, than on tiiem, forming a pit or seam furrow. A 



334 CORROSION OF TEE BEAMS AND ANCHORS. 

number of disastrous failures of important ferric structures have 
occurred from this type of corrosion. 

The quantity of metallic iron in the best refined brands is 99.8 
per cent. In common bar iron it is 98 to 98.3 per cent. In ordinary 
cast iron, about 93.5 per cent, and 2.5 per cent of graphitic carbon. 

The porosity of volume in ordinary cast iron is 1.41 per cent; in 
different kinds of Bessemer steel, 0.41 to 1.20 per cent. In a loco- 
motive tire ingot, 0.57 per cent. For a hard iron tire, 0.97 per cent. 
In a basic-iron rail ingot, 1.95 per cent. In a basic-steel ingot, 1.22 
to 2.17 per cent. 

At an annual convention of the American Institute of Architects, 
in a discussion on the use of iron and steel in the construction of 
modern high buildings, it was reported by one of the leading archi- 
tects in the United States that the iron beams removed by him from 
the old Times building, though in use only thirty-five years, were 
rotten with rust. They were enclosed in eight inches of brickwork, 
forming the arches that supported the pavement over the vault 
where the steam-boilers were placed, and though always dry, yet 
had been exposed to ordinary fire-room vapors. They had been well 
painted with iron-oxide paints and protected from external moisture 
by an asphalt covering. The iron came off in strips, clearly show- 
ing that the rust had followed the lamination of the iron, the web 
of the girders being so rotten as to be easily broken by the fingers. 
Other examples of sidewalk beam corrosion are given on page 269. 

Anchor-stocks made from hammered iron always show less corro- 
sion than cable links. In both cases only the best quality of iron, 
that contains but a small quantity of cinder, is employed. The 
link corrosion is in the form of strips, following the fibre of the metal, 
while the anchor-stock generally corrodes in a compact scale form. 
A different rate of corrosion exists in the cable links at the end, where 
they are welded from that shown in the body. The iron in or near the 
weld is refined and more dense, also has the fibres interlaced; all 
these points have a tendency to delay corrosion, which is more rapid 
where the fibres are undisturbed by the hammer. 

Tunnel Shields and Submarine Metal Corrosion. 

With the recorded instances of the corrosion of iron and steel to 
judge from, it may be pertinent to ask, how long will the metal lining 
of submarine timnels last? Notably, those laid in sea-silt or ocean 
mud. The integrity of the brick lining is wholly dependent upon 



COTtROSlON OF TUNNEL SHIELDS. 336 

the raetal shield put in place as the work progresses, and without 
which the construction of the tunnel would be impossible. Even 
when completed, these constructions have a small margin of strength 
above that necessary to get the metal into place during construc- 
tion, and none at all as a reserve for the loss in strength from the 
inevitable change in the metallic work exposed to sea-water, which 
begins just as soon as the shield is in place; and in the case of tough 



Fio. 48. — CoiTOedon of & st«el plate from the Washington Street r^way bridge 
in Boston. 

close-grained cannon metal, has been ascertained to be at a rate of 
about six inches in one hundred years. 

The lining plates and ribs for tunnel-shields are sddom over 
three inches thick. The different sections of the shield utterly pre- 
clude any material strength to be derived from their circular form, 
made with bolted joints. The bolts holding the shield sections 
together and to each other are relatively small, and will be the first 
to yield to the effects of saline corrosion. 

The brickwork or concrete lining of the tunnel, however thick, 
or the thin coat of partly dried paint, will not protect the shield 
metal, for any appreciable time, from the change to a plumbago-Hke 
substance, which does not require the presence of air to produce it. 
The passage of railway trains through the t\mnel will set up an undu- 
lating or vibratory movement through the tube resting on its bed 
of salt silt, like a log of wood in a mill-pond, being hardly more resist- 



336 



CORROSION IN TUNNELS. 



ant to a change of position from any force, and relatively not a hun- 
dredth part as strong. 

However strong such metallic shields and masonry-lined con- 
structions are when driven through rock or earth, in or through 
salt water or the saturated silt of salt water, they are the most 
treacherous and dangerous of all engineering devices yet conceived, 
affecting the transportation and safety of the public. 

Metallic salts and acids in water intensify the corrosion of all 
metals exposed to their action either by immersion or by condensa- 
tion of the vapors from them. The metal-work of railway tunnels is 
disastrously affected by the condensed vapors of sulphurous and car- 
bonic acid and the moisture due to such locations, the corrosion 
of the metals decreasing the resistance of the water to voltaic circuits; 
this corrosion by liquids being voltaic phenomena in all cases, and 
in many cases is intensified by the moisture being in the form of 
drops instead of being uniformly spread over the whole surface. 

The cut (Fig. 49) from the Railroad GazetUy November 23, 
1894, represents a section of a seventy-six-pound tee-rail laid 
in the Musconetcong Tunnel, removed after being laid 
five years, having lost more weight by corrosion than wear. 
The dotted lines show the original size of the rail, and 
the full lines its present worn 
and corroded size, which is very 
marked. The rails were removed 
on account of their strength having 
been seriously affected by the cor- 
rosion. The tunnel is very damp, 
and a great deal of sand is used 
by the engines, which kept the 
base of the rail covered, the 
vibration caused by the passage of 
the trains having a tendency to 
remove the thin scale of rust almost 
as rapidly as it could be formed. 
There was but little apparent dif- 
ference in the corrosion, whether 
between the cross-ties or where the 

rail rested upon them. The flanges of the wheels removed the rust 
as it formed on the side and top, leaving a clean surface that would 
sensibly corrode between the intervals of the trains. 

In the St. Gothard Tunnel, 49,168 feet long, the air remains 



^ 



-^ 



J 



Fig. 49. 



CORROSION IN TUNNELS. 337 

almost motionless for twelve hours per day, and though the accu- 
mulation of carbonic acid is rapid, and a part of it is absorbed by 
the great quantity of water present, the air is almost unrespirable, 
and causes a great deal of distress to the workmen, and the corrosion 
of all metal-work inside the tunnel is very rapid. 

The water that trickles down the walls of the tunnel is the conden- 
sation of the exhaust gases from the locomotives. It contains sul- 
phuretted hydrogen, sulphur dioxide, ammonia, and carbon dioxide, and 
other combustion gases. The rails are renewed every ten years, and the 
telegraph cables in the tunnel require exceptionally strong wrappings. 

In the new Simplon Tunnel, 64,718 feet long, forced draft is 
proposed, requiring over 500 H. P. at the ventilator shaft. The 
fans render an effective duty of 65 per cent, 1760 cubic feet of air 
per second being required for ventilation. 

In general, the corrosion of metals in tunnels where the rails are 
bedded on cast-iron chairs is represented by cast iron, 100; wrought 
iron, 129; steel, 133. 

In the Arlberg Tunnel, 33,587 feet long, the corrosion of the 
rails and other metals is so rapid, that they all require renewal every 
ten years. The corrosion comes principally from the condensed 
gases of the locomotives, though the traffic is light and a good qual- 
ity of coal is used. The temperature of the tunnel remains almost 
uniformly 75® F. throughout the year. 

The amount of free sulphuric acid in the exhaust gases from tun- 
nel locomotives using a good quality of bituminous coal has been 
found to be about 5 pounds per hour, varying from -0.3 to +7.9 
per cent. 

In a tunnel in France 2850 feet long, where but little water came 
through the walls, the 78-pound-per-yard rails were replaced after 
the passage of 230,000 trains at a speed of nineteen miles per hour. 
They had been eleven and one-half years in service and had lost in 
weight 18i pounds, or 24.166 per cent per yard. The corrosion was 
general over their whole surface, but the rolling action of the wheels 
on the head increased the corrosion at that point by keeping the 
metal bright and removing the rust as fast as it formed. 

Rolled-steel cross-ties for the Indian State Railway, laid in soil 
that was not actively corrosive, at the end of ten years had not 
corroded to any greater extent than the rails laid on them. 

In other locations where steel cross-ties were embedded in soil 
containing sea-sand and saltpetre, both highly corrosive, the rails 
and cross-ties were heated to 300° F. and immersed in a hot bath 



338 



CORROSION OF TEE RAILS ON DOCKS. 



of 3 parts of coal-tar pitch and 1 part of petroleum dead oil. 
This coating was firm and tough, and did not flake or scale off in 
transportation or in the laying of the rails. At the end of twenty 
years the metal had not corroded to any appreciable extent, except 
where mechanically injured. This coating was practically the coal- 
tar dip used on water-pipes by English founders. 

Mr. Otto Herting cites the instance of the corrosion of some 
tee-rails used as girders in a Cape Breton mine, that had been aban- 
doned twenty years. The metal was changed to a grayish-brown 
color, could be cut, and had a specific gravity of only 2.053. The 
metal powdered in a mortar was magnetic. It analyzed as follows: 

Iron 31 . 50 per cent. 

Graphitic carbon 24. 10 

Silicon 14.20 

Manganese 1 . 93 

Sulphur 1.00 

Phosphorus 5.85 

Undetermined and loss 21 .42 



<< 




â– >-kt). 



100.00 

Fig. 50 represents the corrosion of steel rails laid upon docks 
and other places contiguous to sea-water, where the effect of cor- 
rosion was equal to about 4 p?r 
cent each year upon the weight of a 
32-pound rail per yard. The details 
of the rail from which the cut was 
made were contributed by Mr. Del- 
prat, Chief Engineer of the Sumatra 
State Railway, through Mr. J. W. 
§ Post, Divisional Chief Engineer of 
the Netherlands State Railwav. 

The annual report for 1900* of 
the Samarang-Joana Steam Tram- 
way Company (of Java) states 
that a considerable number of steol 
rails had to be removed from the 
harbor tracks on account of cor- 

FiG. 50.— Rail section of the Su- rosion, which amounted to 10.7 kilos 
matra Railway, showing the 
effect of corrosion by sea-water in per meter. 

ten years. Mining metal is exposed to se- 



10 



_ _v 




* Engineering News, November 21, 1901. 



CORROSION IN MINES, THERMO-ELECTRIC ACTION. 339 

vere corrosion from many sources, the presence of sulphurous water 
from decomposed pyrites and other minerals, aided by heat, inten- 
sifying the action. 

Thermo-electric currents arise from changes in temperature and 
set up voltaic action which, though slight and not easily detected, 
will enlarge all fissures, cavities, and seams sufficient to sap the 
strength of the metal. 

Dr. Henry Wurtz* proposes an electro-chemical process for 
protecting mine metal, by connecting all of the fixed metal in the 
mining plant as the negative element, with a dynamo of sufficient 
force to overcome the galvanic energy of the surfaces when exposed 
to the mine's corrosive liquids, the positive terminal to be con- 
nected to a mass of hard coke in the mine sump. These conditions 
vary but slightly from those existing in a ship, and it is not improb- 
able that both systems could be made to work effectively. 

In a number of instances where the whole system of mining-pipes 
required renewal every two years, corrosion was complet€ly stopped 
by coating them wuth an enamel in the following manner: 

The pipes were first pickled in a bath of hydrochloric-acid solu- 
tion to free them from the foundry scale, then washed thoroughly, 
and dried. The pipes then received a coating of 34 parts of silica, 
2 parts of soda, and 15 parts of borax. These were mixed in 
a little water and the pipes exposed for ten or fifteen minutes 
in a dull red-hot retort. A second coating was applied, composed 
of 34 parts of feldspar, 19 parts of silica, 24 parts of borax, 16 parts 
of oxide of tin, 4 parts of fluor spar, 9 parts of soda, and 3 parts of 
saltpetre. These were melted together in a crucible. When cold the 
mass was ground to a fine paste in a little water and applied to the 
pipes with a brush. The pipes were then exposed in a muflSe to a 
white heat. The enamel so formed thoroughly united with the iron, 
and has protected the pipes for over forty years, and they are appar- 
ently in a good condition now. 

Tfie Journal of the Society of Chemical Industry (Ix)ndon), Febru- 
ary 28, 1894, details some experiments upon the galvanic action of 
sea-water upon iron and steel structures in various relations with 
each other, as constructive parts of trusses, boilers, etc., to prevent 
the corrosion. The use of zinc and other easily oxidized metals 
and alloys is suggested, to be so placed and connected to the 

♦"Preservation of Metals from Corrosion by Electric Polarization." Engi- 
neera* Magazine ^ VoL VII, No. 3, May, 1894. 



340 CORROSION OF METALS. 

structure that they will form the electro-positive element of the 
ever-present galvanic circuit, and by their decomposition protect 
the structure, or at least aid the paint coating in its mission of pro- 
tection. 

These protective features, proposed for the internal parts of a 
ship, do not apply to the protection of the external surfaces, where 
an entirely new set of conditions are in force, owing to the nimier- 
ous rivets employed to hold the plates together and to the frames, 
and which are necessarily unprotected from the many sources of 
corrosion herein mentioned. 

Professor V. B. Lewes * of the Royal Naval CoUege, Greenwich, 
England, at a recent meeting of the Institute of Naval Architects, 
London, states: 

"The rusting of iron and steel is a definite chemical process, 
due to the conjoint action of air, moisture, and carbon dioxide upon 
the metal. The increased rate of chemical corrosive action due to 
a local increase of temperature is noticeable, and may be due to 
galvanic action set up between portions of the same metal at differ- 
ent temperatures. 

"It is an undoubted fact that the double bottom of ship plates 
near the boilers corrodes more rapidly than similar plates in other 
parts of the vessel, and the increase in temperature near the boiler 
is the only factor. 

"It is also noteworthy that the plates at the bottom of the cellular 
spaces which are kept cool by contact with the sea-water do not 
corrode; and cases are noted in which parts of a plate, which get 
locally warmer than other parts — ^although the difference can only be 
a few degrees — corrode much more rapidly than the cooler portions. 

"Experiments show that the rapid corrosion foimd in the double 
bottoms near the boilers or other sources of heat, is due to galvanic 
action, and not to the increased chemical activity due simply to 
the increase of temperature. As the ashes are drawn and quenched 
with sea-water near these exposed plates, no doubt some of the corro- 
sion can be traced to the gases thus formed; the sulphur in the ashes 
also contributing its effect." 

Mr. William Thomson, F.R.S., read a paper on "The Influence 

* A paper read at the thirtieth session of the Institution of Naval Architects 
by Prof. Vivian B. Lewes, F.R.S., F.I.C., Royal Naval College Associate, April 
12, 1889; and published in full, Scientific American Supplementf Vol XX\^n, 
No. 709, August 3, 1889; pp. 11, 320. 



CORROSION OF METALS IN CONTACT. 341 

of Some Chemical Agents in Producing Injury to Iron and Steel," 
before the Manchester Association of Engineers, November 25, 1893, 
in which he refers to the interesting and exhaustive experiments 
made by Mr. Thomas Andrews, F.R.S., on the galvanic action which 
takes place between iron and steel, and between iron of different 
kinds and steel of different kinds, viz.: 

"The galvanic action between wrought iron, cast metals, and 
various steels during long exposures in sea-water." Institute of 
Gvil Engineers, Vol. 1883-84, Part III. 

"Corrosion of metals during long exposures in sea-water." Insti- 
tute of Gvil Engineers, Vol. LXXXII, 1884-85, Part IV. 

• "The relative electro-chemical positions of wrought iron, steel, 
cast metals, etc., in sea-water and other solutions." Royal Society 
of Edinburgh, Vols. 1883-1889. 

Mr. David Phillips's paper. Institute of Marine Engineers, 1890. 

In the above-named articles Mr. Andrews shows that while some 
varieties of iron and steel remain constantly electro-positive or electro- 
negative to each other, others change, taking opposite positions toward 
each other, while others again change positions constantly during long 
periods, these changes always producing rust. 

" It can be easily understood that while there is no material voltaic 
action between two pieces of steel or two pieces of iron, or of pieces 
of steel and iron, there may be conditions on the surface of one plate 
or rivet which may act strongly as an electro-negative element, and 
produce rusting on the metal in contact with it. A piece of iron 
immersed in weak nitric acid begins to dissolve at once. A similar 
piece placed in strong nitric acid, touching it for a few minutes with 
a piece of platinum wire, and then putting it into the weak nitric 
acid, will not dissolve, it having been rendered passive; and simi- 
larly, there is reason why one piece of iron may act electro-nega- 
tively toward another piece of the same metal, on account of some 
slight alteration of its physical properties, by hanunering, such as 
closing the riveted seams of plates, calking seams, setting tubes, 
etc., or it may have attached to it some oxide of iron, which always 
acts electro-negatively toward any metal with which it is in contact, 
and induces oxidation in such metal." 

The commission of English engineers,* appointed by the English 

♦Transactions Institution of Marine Engineers (English), May 13, 1890. 
Minut€8 of Proceedings avil Engineers (English), Vol LXXVII, p. 323, and Vol 
LXXXII, p. 281. 

"Electro-chemical Effects on Magnetizing Iron." Proceedings Royal 



842 CORROSION OF METALS IN CONTACT. 

Government to investigate the cause of the failure of the Tay Bridge, 
reported that where cast iron and wrought iron were connected 
by rivets in many parts of the same structure (as they were in this 
one), the rivets and connecting wrought-iron work, where connected 
to the cast-iron members of the structure (columns, flanges, span- 
drels, etc.), had corroded to such an extent as to be below the point 
of stability by the local galvanic circles formed at numerous points 
in the structure where the two metals were in contact, and the corro- 
sion thus established was the cause of the disaster. 

Mr. St. John Day, in a paper read before the Institution of Engi- 
neers and Shipbuilders, Scotland, February, 1880, stated that " the 
li-inch diameter bolts holding the ties to the piles on the Tay Bridge 
were so corroded that they would have to be replaced every four to 
six years. Some of the bolts were found to be corroded away to 
one-half of their o ginal size." 

The Scotland Board of Trade now prohibits the connection of 
cas1>-iron columns with wrought-iron columns or ties. 

The corrosive action noticed in the riveted sections of the tubular 
bridge over the St. Lawrence River at Montreal, Canada, referred 
to before, resembling Fig. 39, Chapter XXVIII, was of a similar 
nature to the above example. In this case, while the corrosion was 
of almost unparalleled amount and virility in the whole structure^ 
the rivets that held the floor-beams and track-stringers in place^ 
and were under the greatest strain and subject to vibration and 
shock from the passing of the railway trains, were corroded the most^ 
though all of these parts were of wrought iron to wrought iron, but 
varied in quality from common iron to refined iron. 

An important question presents itself to boiler-makere : whether 
it is safe to rivet steel plates to iron plates in steam-boilers, or even 
in other constructions, particularly where exposed to high tempera- 



Society, Vol. Xni, p. 429; Vol. XUV, p. 152; Vol. XLIV, p. 176; and Vol 

LH, p. 114. 

"On the Corrosion of Metals in Sea-water." Minutes of Proceedings Institu- 
tion of Civil Engineers (English), Vol. XI.XVH, p. 323, and Vol LXXII, p. 281. 

"The Action of Tidal Streams on Metiils." Proceedings Federated Institu- 
tion of Marine Engineers, Vol. I, p. 191. 1890. 

Report of the meeting of the British Association for the Advancement of 
Science, Edinburgh, 1892. 

"The Wasting and Protection of Iron in Sea-water." 

From "Notes on Docks and Dock Construction," by C. Colson, M. Inst. C.E. 

The Practical Engineer, London, October 19, 1893. Vol. X, No. 399. 



CORROSION OF METALS IN CONTACT. 343 

tures or frequent changes of moderate temperatures, or to use iron 
rivets for steel plates or steel rivets for iron plates? 

Cases are shown where furnace plates of steel riveted together 
with iron rivets are badly rusted or pitted in the vicinity of the 
rivets while the latter remain intact. To determine how iron 
stands to steel and how different amples of steel stand to each 
other, Mr. Thomson made an extended series of experiments, using 
a Thomson's tangent galvanometer to measure the electrical cur- 
rents generated in the corrosion of iron and steel, both singly and in 
connection with each other, and when immersed in different fluids, viz. : 
sulphuric acid (one part to nine of water), caustic-potash solution 
(specific gravity, 1.311), and chloride-of-ammonium solution (specific 
gravity, 1.033), the latter representing electrically the ordinary con- 
centrated water found in steam-boilers. 

The details of the experiments are important, but I wiU give 
only the results obtained, viz. : 

" When an iron rivet and a piece of the above-mentioned corroded 
steel furnace plate were placed in contact and inunersed in the weak 
sulphuric-acid bath, at first the steel was electro-negative to the 
iron, but in a few moments it changed, and afterward the iron was 
electro-negative to the steel. When placed in the chloride-of-ammo- 
nium solution, at first the iron was strongly electro-positive to the 
steel, and afterward became wecMy electro-negative. When placed 
in the caustic-potash solution, the steel was strongly electro-positive, 
but the current gradually became weaker and weaker until it practi- 
cally ceased. A new steel rivet in an iron plate, a steel rivet closed 
by a machine and held until nearly cold, an iron rivet closed on a mild 
unworked steel plate, all reacted strongly among themselves. The 
iron when first brought into voltaic contact with the steel was strongly 
electro-positive to the steel, being presumably strongly acted upon 
by the solution, but after a few minutes almost ceased action or 
became reversed; and so far as the tests demonstrated as a whole, 
it was to the effect that it was quite as safe to bring iron and steel 
in close mechanical contact with each other as two different kinds 
of steel or two kinds of iron. Corrosion was developed in some 
degree in the contact of all different metals to each other." 

It is cited that a number of torpedo boats of the French Navy, 
that had been constructed within ten years, and that had not made a 
thousand knots of sea service, were found to be so corroded at the 
water-line, though well painted from the first with anti-corrosive 



344 CORROSION OF METALS. USE OF METALLIC-SALT PAINTS. 

paints, they had to be condemned for service; while other boats of 
the same class that had never been in commission, but had been 
laid up under cover, had, as the report says, "eaten their own heads 
off by corrosion," and were condemned for the same cause. In these 
cases the corrosion had been in progress under the paint covering, and 
showed but little sign of its extent or progress, imtil the plates were 
so corroded in spots, many of them of large area, that the hammer 
used in testing the plates broke through the skin of the boats under 
the effect of blows that would not drive a nail into a pine block. 

The use of anti-corrosive or anti-fouling paints, containing salts 
of any metal, is attended with the greatest danger to the coated 
structure. These pigments are extremely sensitive to the presence 
of saline elements in moisture, their action being to rapidly dissolve 
portions of the iron, and to deposit the metal which they contain 
upon the surface of the plates, and these deposits exciting energetic 
galvanic action, cause corrosion and pitting to go on with alarming 
rapidity. 

Both mercury and copper salts are offenders in this way, but 
copper is by far the more objectionable, from the fact that the salts 
formed by the action of the sea-water upon the compounds used in 
the compositions are far more soluble than the corresponding salts 
of mercury, and are therefore liable to be present in much larger 
quantity, and so exert comparatively a much more injurious action 
on the plates. 

As an illustration of this, two equal portions of sea-water were 
saturated, the one with copper chloride, the other with mercuric 
chloride, and into each a piece of steel planed upon one side and of 
about equal weight and size, was placed and left for four days. At 
the end of this period the two plates were removed, and after being 
cleaned and dried, were again weighed, when it was found that the one 
exposed to the copper-saturated sea-water had lost 22.2 per cent in 
weight, while the plate exposed to the mercurial solution had only 
lost 3.6 per cent, this being due to the much larger amount of the 
copper salt soluble in the sea-water. 

On placing these plates in clean sea-water, corrosion went on in 
each case with extreme rapidity, and after being exposed for a month, 
they had both wasted to about the same extent; that is to say, when 
once deposited on the iron, mercury is practically as injurious as 
copper. See further data in Chapter XXXV. 

In the year 1835, Mr. Peacock tried zinc plates on the bottom 



CORROSION OF METALS. USE OF ZINC. 346 

of H.M.S. Medea, and in 1867 Mr. T. B. Daft again brought the sub- 
ject forward, Sir Nathaniel Bamabay, Mr. Mclntyre, and others 
also suggesting various plans of attachment. In 1888 Mr. C. F. 
Henwood read a paper before the United Service Institute, strongly 
advocating zinc sheeting as attached by his system. 

When the galvanic contact was small, then the sheeting had a 
certain life, but afforded but little protection to the iron, and gradu- 
ally decayed away in a very uneven fashion; while in those cases 
where galvanic contact was successfully made, the ship on several 
occasions returned from her voyage minus a considerable portion 
of her sheeting. 

Another drawback to the use of zinc sheathing is one which was 
found when it was used to coat wooden ships, and that is that sheet 
zinc, like every other metal, is by no means homogeneous, and that 
for this reason the action of the sea-water upo it, leaving out of 
consideration galvanic action, is very unevenly carried on, the sheet- 
ing showing a strong tendency to be eaten away in patches, while 
the metal itself undergoes some physical change and rapidly becomes 
britt e. 

Attempts have been made to galvanize the iron before the building 
of the ship, but Mr. Mallett showed, in 1843, that this coating was 
useless when exposed to sea-water, as in from two to three months 
the whole of the zinc coating was converted into chloride and oxide; 
and when, therefore, galvanizing is used care must be taken to pro- 
tect the thin coating of zinc. In any case the galvanizing must be 
done after the plates are riveted up, as any break in the surface 
would set up a rapid wasting away of the zinc, and the process could, 
therefore, be only used on small craft. Fresh water has less action 
upon the zinc than sea-water, and for this service galvanizing would 
be attended with some measure of success, the rapid wasting of the 
zinc in sea-water being due to the salts. 

As has been before stated, if plates of iron or steel and one of 
copper be joined together or placed in conmiunication and immersed 
in sea-water, acidulated solutions of water, or of mineral salts or 
oxides, the ferric body becomes electro-positive to the copper and 
is rapidly corroded. The corroded metal is always found in com- 
munication with the positive pole or current of electricity, the fluid 
soon becoming red from the rust formed. 

In the corrosion of marine boilers, contact of different metals, 
strain, heat, and chemical action from the sea-water are all present 



346 CORROSION OF BOILERS. USE OF ZINC TO PREVENT. 

and acting towards the same end. They are all of different ix>ten- 
tial and electro-positive, and none counteract each other, but all 
attack the boiler metals. 

The voltage in this case is distinctly recognizable and evidently 
much different from the instances cited by Mr. Thompson of metals 
under strain only. In one reported case, it was one ohm, corrosion 
was marked and the current grew stronger as the corrosion increased. 

Fresh water and solutions other than sea-water, also vapors, are 
corrosive agents to boilers, the corrosion of which is modified, corrected 
and rendered nil, by the use of electrogens, or heavy cast-zinc 
plates. In old boilers using fresh or salt water, the corrosion in 
progress is arrested by the use of the electrogens, so long as any 
appreciable amoimt of zinc is present. When the zinc is wasted 
away or removed, "bleeding" from the boilers at once begins, par- 
ticularly in old boilers or in the tube settings. 

In marine boilers, zinc 10" X 6" XI" to the amount of one square 
foot of surface to 50 square feet of heating surface is placed in clean, 
firm, metallic contact with the internal steam or water surfaces. 
Too much zinc is hardly possible and is better than too little. The 
amount of zinc can be reduced after a time to 75 or 100 square feet. 
The zinc must be placed in absolute contact with the bright metal 
at a number of points. Suspending the zinc in any form in trays 
or baskets will not prove effective. Zinc is slowly dissolved in hot 
water, and deposited as a sediment that can be removed by the blow- 
off, carrying with it any old scale or rust loosened by the galvanic 
action of the zinc. The boiler fluid contains a white flocculent pre- 
cipitate of zinc (zinc oxide). If the water contains the sulphates or 
carbonates of lime or magnesia, sihcates or other minerals, that form 
the usual hard, vitreous scale, the precipitated zinc oxide unites with 
them and holds them in solution until blown out. 

Zinc causing old boilers to bleed, might be considered an injury 
instead of a blessing. It indicates that the boiler needs repairs to 
prevent future disasters. 

The amount of zinc in boilers for land service using waters con- 
taining mineral substances, has not been so clearly ascertained as 
in the case of marine work; but the results in the latter case are a 
good basis to reckon from. 

Any neutral salt in water which decreases its resistance, will 
enable it to act as the necessary liquid medium in a voltaic circuit. 

The disintegration of the zinc in boilers forms the same oxide 



CORROSION OF BOILERS. USE OF ZINC TO PREVENT. 347 

that is formed in the roasting furnace for pigments, i.e., 80.344 parts 
of zinc and 19.656 parts of oxygen. The hydrogen set free replaces 
that lost in the heating of the water, that in a measure is broken up 
at all heats below a low red heat, where complete dissociation of the 
hydrogen occurs. In both the steam and water, the flocculent par- 
ticles of the zinc readily unite with any ammoniacal, carbonic, or sul- 
phuric acids, saccharine, or other organic vapors or liquids present 
to form sulphates, carbonates of zinc, etc., that would act mechan- 
ically in the water to prevent deposit and cause corrosion. 

Central-heating-system pipes develop a virulent corrosion. The 
reason for some of the cases is difficult to find. This corrosion has 
caused the abandonment of the pipes returning the condensed water 
to the boilers, and in some cases the failure of the whole system. 
The corrosion occurs in both the steam-pipes and water-return pipes, 
being more marked in the latter. In screwed-end pipes the corro- 
sion first attacks the heel of the pipe threads as zones of disintegra- 
tion and extends until the whole pipe is affected, though no corrosion 
except at the joints may be noticed. 

In the steam-pipes corrosion takes the form of pin-holing or pitting, 
from the inside at any part of the pipe and does not develop into a 
general corrosion of the surface as in the case of the hot-water return 
pipes. Whenever a blow-out or pin-hole from corrosion occurs in 
the steam pipe, a closing down of the metal around the hole by a 
peen hammer stops the leak, which seldom reopens. Peening the 
metal has made it more dense and ess liable to corrode. 

Cast-iron steam-pipes are less affected than wTought-iron pipes, 
but the joints draw badly, owing to the temperature changes that 
cause a leakage that frequent caulking only momentarily corrects. 
Cement joints are unreliable under high temperatures, while rust 
joints owing to their own corrosion burst the sockets of the pipes. 

In central-heating systems, the losses from leakage and condensa- 
tion amount to from 30 to 35 per cent yearly, aside from the loss of 
the return water; while the corrosion losses are about 10 per cent 
of the cost of the pipe lines. 

Particles of dirt, cinder in excess, unabsorbed carbon or manga- 



"I^se of Zinc in the Steamship Hindostan," Engineering, Auji^st 7, 1878. 

"The Corrosion of Steamship Boilers." The Practical Engineer, Vol. X, 
Sept«rnber 28, 1894. 

"The Corrosion of Boilers." The Engineer (London), Vol. LXXVIII, 1894, 
p. 208-281. 



348 CORROSION IN CENTRAL HEATING SYSTEMS. 

nese and impurities in the metal have all been blamed for the erratic 
disintegration of the pipes, which continues after many of the above 
causes have been removed. 

In all of these pipes a low density, open-grained filament-formed 
iron, cinder in excess, heat, motion of vapor and water under pres- 
sure are all present, and no protective covering of moment. The 
hot water is more eflfective than steam in keeping the pipes clean 
and bright, ready for corrosive influences. 

The disturbance of water by high heat in being partially disso- 
ciated has been already explained in this chapter. The action of 
zinc on e.ectrogens, in the case of marine boilers and sea-water 
corrosion, is always favorable for the preservation of the metal. The 
use of zinc to prevent corrosion in steam-pipes, radiator and heating 
lines, could be hardly less favorable. 

M. Loudin's (Comptes Rendus) experiments on the corrosion of 
iron immersed in water usually found in steam-boilers, was: That 
with both ordinary and distilled water, the temperature had a very 
important influence, viz. : 

" At 68° F. the quantities of oxygen absorbed per square foot of 
iron surface per hour, when immersed in distilled water was 0.258 
grain and in calcareous water 0.330 grain. At 212° F. the quantities 
rose to 2.364 and 2.579 grains. The immersion of iron in water at all 
ordinary temperatures was attended by the evolution of hydrogen, 
the action being the least in distilled water. At a temperature of 
260° F., the decomposition of distilled water was equal to the absorpK 
tion of 0.01 grain per square foot of ferric surface per hour, and for 
calcareous water, 0.0129 grain. For water containing one-fifth part 
of cr3rstallized chloride of magnesium, corrosion was 0.0182 grain per 
square foot, and for water saturated with chloride of sodium, 0.05 
grain ; for sea-water of usual density, 0.067 grain, all per square foot 
of iron surface per hour immersion." 

Corrosion Increased by Stress, 

The tendency of iron to change its phj'^ical properties by a change 
in the condition under which it may be placed in ordinary structural 
work is strikingly shown by the following instance taken from Engi- 
neering, April 27, 1894, and reported by Mr. Oswald Brown, M.I.C.E., 
of 32 Victoria Street, Westminster. 

''The cut (Fig. 51) shows portions of the bar dark and corroded, 
while the intermediate layers have remained bright. The bands of 



CORROSION IXCREASED BY STRESS. 349 

ruBt extend over both ends of the bar, ijving it the appearance of 
being built up of layers of two different metals. The bar, wliich is 
of the best Yorkshire iron, gave under test the following results: 

"Tensile strength, 54,230 pounds per square inch; elongation on 
8 inches, 28,4 per cent; contraction of area, 49.6 per cent. No traces 
of laminatJoD were shown during the test, but some months after, the 
bar was found in the condition illustrated, which shows that it con- 



Fia. 51, — Effect of strain on the cornvdon of titHL 

msts of layers of different chemical composition, those which have 
rusted being electixj-positive to the other portions of the bar." 

Iron rivetfl and iron plates in some ca.sea show the rivets corroded 
and the plates unaffected, and Bometimes the contrary, and so with 
steel rivets and steel plates; also iron rivets in steel plates or steel 
rivets in iron plates all show the most erratic evidences as regards 
corrosion, in many cases without reference to the character of the 
water used in the boiler or to the external conditions. As a rule, all 
analyses of the plates, rivets, and other material used in boiler work 
are made from samples as they come from the manufacturer's hands, 
and before being worked. Hence, when corrosion of either plate 
or rivet has attracted attention, it is seldom possible to get a sam- 
ple of that particular make and lot of rivets to analyze to show 
what phyacal changes were developed by the processes of 
heating, closing the rivet, cold-hammering the head, chipping, 



350 CORROSION INCREASED BY STRESS, 

caulking, etc. These processes, abo punching instead of drilling 
the holes, develop corrosion, that takes the form of pitting around the 
rivets and furrowing on the sheet joints. 

Mr. Thomas Andrews, F.R.S., reporting to the British Institution 
of Civil Engineers, states his conclusions On the Effect of Stress on the 
Corrosion of Metals.* In brief they are : 

"That wrought iron and various steels, when exposed separately, 
without liability to galvanic action other than local, under the action 
of sea-water for long periods, showed a greater corrosion on the part 
of all the steels than the wrought iron; the advantage in favor of 
the iron compared with the steels amounting to 25 per cent and 
upward. It was also noticed that corrosion was increased in the 
steels in proportion as the percentage of combined carbon was greater. 

" It was found that the galvanic action between wrought iron 
and steels induced a largely increased corrosion in both metals. 
It was also found that the upper and lower portions of a metal struc- 
ture, or vessel, although composed throughout of the same metal, 
were exposed to electrolytic disintegration from the galvanic action 
set up by solutions of different salinity on the metal; conditions 
found almost constant in tidal streams, brought about by the gradual 
rise and inflow of salt water and the outward flow of fresh water; 
and there are strong evidences to show that magnetic influence tends 
to increase the corrosion of metals. 

*' When, however, the strained metal is in galvanic circuit or com- 
bination with the unstrained metal in any solution, an increased total 
corrosion ensues from the galvanic action, which research has shown 
to arise consequent on the different potential between the two. 

" It was demonstrated that stress of any kind considerably alters the 
physical properties of both iron and steel, by increasing their rigidity 
and rendering the metals harder, also greatly reducing their prop- 
erties of elongation or ductility. It requires a higher tonnage to break 
a strained than an unstrained bar of the same metal. A tensile stress 
applied to a wrought-iron shaft, that produces an elongation of only 
2 per cent, increases the tensile resistance of the metal 2.66 percent. 

♦ Proceedings of the Institution of Civil Engineers (English), Vol CXVin, 
1893-94, Part IV, p. 356. 

The Practical Engineer (London), Vol X, No. 398, October 12, 1894. 

Iron Age, Vol. No. 17, October 25, 1894. 

Minutes of Proceedings Institute Civil Engineers, Vol LXXXVII, p, 840, 
and Vol XCIV. p. 180; also Vol CV, p. 161. 



CORROSION INCREASED BY STRESS. 351 

"From the observations it was manifest that the stresses applied 
to metals altered their structure, rendered them harder in nature, 
and more liable while in their strained condition to be acted upon by 
sea-water, or other waters, than in their ordinary normal or softer 
condition. The experiments, however, indicate that an increased 
total corrosion, in excess of the normal corrosibility of the metal, 
occura in a metallic structure, from the action of the local galvanic 
currents which are shown to be induced between strained and un- 
strained portions of the same piece of iron or steel forging, bar, or 
plate. Hence a strain occurring in a metallic structure tends, omng 
to the local galvanic action thus set up, to increase any corrosive 
forces which may be deteriorating the metal of which it is composed." 

The details of the experiments are: Pieces of iron and mild steel 
of known character were submitted to tension, torsion, and flexure 
strains, to ascertain the changes made in the metal, and if corrosive 
effects were in any manner due to stress. For tension, a bar was 
strained in a testing machine until an elongation was produced of 
23 per cent in three inches, and at the point of reduced area the bar 
wai cut in two. 

The halves were then turned down at the shackle or vise end, 
where they had been subjected to little or no stress, until they had an 
area equal to the end half at the point where contraction of area had 
occurred, both pieces being finished exactly aUke and each piece 
represented a section of strained and unstrained metal. They were 
then placed at the same depth in a saturated solution of common salt to 
approximate the action of sea-water on metal, the immersed ends 
representing strained and unstrained metal. An electrical contact 
made between the two pieces of metal, through the medium of a 
delicate galvanometer (Thomson's), the difference in potential or 
corrosibility could be observed. It was found that in each case 
a sensible current was set up between the two halves of the specimen ; 
the strained portion was in every case found to be the electro-posi- 
tive element of the pair, corresponding to the zinc in a galvanic 
couple, indicating clearly that the strained metal was acted upon 
more rapidly by the solution, and more easily corroded than the 
unstrained metal. 

The test made with specimens after being submitted to torsional 
stress » representing a bar that had been twisted through an angle equal 
to half a revolution, and prepared similar to those in the tensile test, 
showed results identical with the tensile strains. In every instance 



352 CORROSION INCREASED BY STRESS. 

the strained metal was the electro-positive element, and was cor- 
roded more rapidly by the sea-water. 

This conclusion was further supported by tests made with iron 
and steel plates, when a flat piece was compared with one bent into 
an U or semi-circular trough; the bent plate in each case pro\'ing 
to be the one most easily acted upon by the solution. 

The experiments throw an interesting light on a subject which 
has hardly received the attention it deserves, and helps to explain 
some of the peculiarities in connection with the wasting of certain 
structures that have been involved in considerable mystery. The 
metals operated upon by Mr. Andrews were large, rolled wrought- 
iron bars and hammered wrought-iron shafts; Bessemer steel and 
Siemens steel forged shafts, also arge bars of soft and hard Bessemer 
and Siemens steel; soft and hard cast steel, and steels made from 
each of the metals aluminum, nickel, silicon, and copper. Experi- 
ments were also made on rolled plates of wrought iron, soft Bessemer 
and soft and hard Siemens steel and soft cast iron. The chemical 
compositions and general physical properties, etc., of all the metals 
are given and tabulated. All the metals experimented upon were 
perfectly bright. 

General results: The average electromotive force obtained be- 
tween strained and unstrained portions of the same metal were, viz. : 

Wrought-iron forged shafts .016 volts. 

Soft Bessemer steel forged 0.019 " 

Hard " " " 0.006 " 

Soft cast steel 0.003 " 

Hard " " 0.003 " 

Silicon steel 0.004 " 

Aluminum steel .004 " 

Nickel steel 0.003 " 

Rolled wrought-iron bars 0.002 " 

Soft Siemens steel 0.005 " 

Hard " " 0.005 " 

Copper steel 0.006 " 

Chromium steel 0.001 " 

Bessemer steel hammered forgings 0.011 " 

Siemens steel " " 0.006 " 

With cold-drawn small steel rods in galvanic circuit with copper 
rods, similar results were noted, the electromotive force between 
strained and unstrained aluminum steel being 0.022 volts, and 
strained and unstrained cast steel being 0.023 volts. 



CORROSION INCREASED BY STRESS, 



353 



In all these tests the strained metal was the electro-positive. 
In the torsional tests the electromotive force was notably higher 
than in the tensile, also in the flexure, tests. 

These electric measurements ought, perhaps, to be regarded as 
tentative indications, establishing a general principle, rather than 
as an absolute measurement for the purpose of accurate comparison 
of the behavior of the various metals. The chemical analysis of 
all the metals was made prior to straining them. These experiments 
extended from a few seconds to over ten days, in which it was ob- 
served that the diflference in the electromotive force between 
strained and imstrained metal steadily decUned from the initial 
amount, but was in no case extinguished. 

Corrosion, or the oxidation of substances by chemical action is 
always accompanied by electrical energy, that may be of more or less 
intensity, or electromotive force according to the substance con- 
sumed. 

Chemical action is probably due to the unbalanced attraction 
among the various molecules of matter lying in juxtaposition, the 
rearrangement of which caused by strain or a change in the thermal 
or electrical conditions of one atom changes them all. It is known 
that a change in either the thermal or electrical conditions develops 
corrosion in certain circumstances but does not in many other cases 
of apparently the same nature. The amount of electromotive force 
developed in the oxidation of a few substances is indicated in the fol- 
owing instances : * 



Substanoes. 



Carbon 

Silver 

Copper. 

Lead 

Iron 

Zinc 

Peroxide of lead 



Heat of Oxidation of 

Equivalent. 
Calories, fi. T. Units. 



2,000 
9,000 
18,760 
25,100 
34,120 
42,700 
12,500 



7,938 
31,742 
74,057 
99,616 
135,415 
169,074 
48,022 



E.M.F. Relative 
to Oxygen. 



0.09 
0.39 
0.80 
1.12 
1.55 
1.83 
0.52 



E.M.F. Relative 
to Zinc. 



-1.74 
-1.44 
-1.08 
-0.71 
-0.28 
0.00 
-2.35 



The corrosion of ferric bodies results from the decomposition 
of water or air by electrical energy. 

As detailed in Chapter III, atmospheric moisture in the presence 
of iron at a temperature of 900® F. releases oxygen and forms the 



♦Thompson's Electricity and Magnetism. 



364 CORROSION OP METAL. ELECTRO-CHEMICAL ACTION. 

black magnetic or stable oxide of iron, that in manufactured articles 
is represented by the Bower-Barflf products. 

Ever}' pound of iron oxide represents the energy of 1.668 pound 
of coal required for its formation. This rust requires .3375 pound 
of water to furnish the necessary oxygen. A pound of iron oxide 
represents the corrosion of 13.13 square feet of metallic iron -jVirv inch 
in thickness and has developed 163^33 B. T. Units (.0641 H.P.), 



Ka.CU 



JIB^ Cl« 




Fig. 52. — Diagram of the coirodbility of metals. 

or equal to an electromotive force of 1.55, to neutralize which would 
require the oxidation of .847 pound of zinc. 

In aU of these oxidations the thermal manifestations are subordi- 
nate, and, with the electrical energy being of low potential, when they 
extend over any considerable period, are unrecognized or difficult to 
determine. 

The affinity of oxygen for hydrogen represents an electromotive 
force of 1.47 volt. The decomposition of water acidulated by sul- 
phuric acid jdelds at the cathode 11.12 parts of hydrogen and at the 
anode 88.18 parts of oxygen; about 1 per cent of ozone bring formed 
from the oxygen. 

The decomposition of 1 pound of zinc for the protection from 
corrosion of marine boilers or other like ferric bodies evolves 9840 cubic 
inches of hydrogen, equal to 5.694 cubic feet, that weighs 210.29 
grains, or .3329 pound. 

The late Henry Morton, Ph.D.,* stated that a pound of zinc con- 
sumed in the following-named batteries develops: 

* "The Maximum Possible Efficiency of Galvanic Batteries." Cassier^s 
Magazine f June, 1895, p. 130. 



CORROSION OF METAL. ELECTRO-CHEMICAL ACTION. 356 



Smee's battery 

Daniels's (sulphate of copper) battery. 
Grove's or Bunsen's (nitric acid) ** 
Peggendorf's (chromic acid) ** 

Sulphuric acid (1 in 9 parts of water). . 



B. T. Units. 



900 
1419 
2722.4 
2827 . 5 
3006 



Horse-power. 



0.35 
0.55 
1.06 
1.14 
1.17 



All losses from resistance being excluded. 

A series of experiments covering several years, upon the corrosi- 
bility of metals, has been made at the University of Wisconsin, under 
the supervision of Prof. Dugald C. Jackson. 

Prof. Jackson, in the discussion of paper No. 901, "Protection 
of Ferric Structures," * referred to the results of his experiments, 
from which I briefly quote: "When a piece of iron or steel is placed 
in a testing machine and its electrical condition is followed up during 
the straining test, its corrosiMlUy appears to increase practically in. 
proportion with the strain, so that a diagram plotted with stress along 
one coordinate and corrosibility along the other appears to be of 
almost exactly similar character to a diagram plotted with stress 
and strain along the two axes of coordinates. 

" Two illustrations of these diagrams are presented, Figs. 53 and 
64, from test pieces of wrought iron 

" In the case of cast iron Fig. 55 shows the stress-corrosibility 
diagrams for two specimens in tension. A comparison of these dia- 
grams with those for wrought iron in tension, illustrated in Figs. 53 
and 54, shows the marked difference between the two metals. Fig. 56 
shows a stress-corrosibility diagram for cast iron in compression. 
The exact forms of the diagrams taken from cast iron depend in some 
degree on the physical character of the specimens, but the diagrams 
shown are typical ones. The effect of strain is small in the case of cast 
iron. 

" The corrosibility of the specimens was measured by determining 
the electromotive forces of the test pieces toward a standard electrode 
in a normal solution. 

"The results of the tests show that in bridge members and 
similar pieces that have been worked, the metal appears to be eoMly 
affected by corrosion, this corrosion being properly characterized as 



♦Transactions American Society Mechanical Engineers, VoL XXII, Paper 
No. 901, May, 1901. 



356 



STRESS CORROSION DIAGRAMS. 



15000 

10000 

BOOO 





19000 

10000 
6000 













VrROU(LHT nON(WeDi/METE 


:r) 








































•9 












nt 


• ' 


^^-^ 
















1 




/ 






8TR 


^S8-S 


^RAl^ 


DIAG 


RAM 
















/ 








Elongatloi^ In In 


shes. 
















) 





5 


1 





1 


1 1 

5 


2 





2 


5 


8.0 1 
















— 


— 










__^ 




f 






/^ 






'^ 














a 






f 




STRI 


88-C< 


>RR01 


ilBILI' 


Y Dl ^QRAM 












A 


1 


Coi 


Toaib! 


iltylo 


elect) 


o-chei 


nlcal 


inlts 












-0.( 


22 


-0.( 


12 


-0. 


X» 


«0. 


X)8 


-»0. 


)18 


•»0. 


)28 


t 1 



Fig. 53. — Stress-strain and stress corrosibility of wrought iron. 













y 


ROUC 


HT 1 


tON ( 


HeDI/ 


METI 


:r) 










IBOOO 

uooo 

BOOO 
































f 




i^ 





























£ 




f 






STRI 


:ss-8i 


-RAIN 


DIAG 


RAM 


















/ 




E 


iongai 


ion in 


Inohc 


B. 





















1 







5 


1 





1 


5 


2 





2 


5 






19000 

10000 

6000 
















— = 
















9 






/^ 




^ 
























r 




8TR 


ES8-C 


ORRC 


SIBIL 


ITY C 


lAGRi 


^M 












^ 


I 


1 


>>ITOI 


iblllt: 


' f n ar, 


Mtrar 


ronll 


1 















c 


, 




>• 


( 

i 


L 


% 


L 


1 


'• 


1 


>• 







Fig. 54. — Comparison of stress-strain and stress-corrosibility diagrams, wrought 

iron. 



STRESS-CORROSION DIAGRAMS. 



357 



caused by electrolysis ; that is, the strained metal is- really eaten 
away and the unstrained metal is not. 

** The experiments give a satisfactory explanation of much of the 
so-called grooving in boilers and corrosion of a similar character. Here 
the strained metal of a punched boiler plate that is not completely 
covered by a rivet-head becomes eaten away. Or perhaps a plate 
becomes strained at a joint by temperature stresses and the strained 



90000 












/ 












i 


f 




IdOOO 










f 






• 














13000 


9 






'/ 








a 






/ 








8000 








/- 








t3 




"V 


r 








4000 






"/ 














/ 














y 













-<).0BBO -O.U582 >U.OEF74 -O.0S68 

OofTOdblllty In arbitrary imlts. 



30000 
















/ 














; 


' 


lonoo 














/ 




• 












/ 




laooo 


a 






( 


^ 








i 






\ 


X 


Si* 






8000 


« 










) 






P 








^ 


/ 






4000 








-^ 












/ 
















y 















•O.0G00 -O.0003 ^.0084 -QJ»46 

Oorroolbllltj In arbitrary unltg. 



Fig. 55. — Stress-corrosibility diagrams for cast iron. 

streak is corroded. In each case the strained metal is of greater corrosv- 
bilityy and it acts as one of the plates of an electric battery in which 
the other plate of the battery is the unstrained metal of the boiler 
shell, and the electrolyte is the water within the boiler. The strained 
metal is the electrode which corresponds to the zinc of the ordinary 
voltaic cell, and it is eaten away. 

" Another illustration of corrosion of this character is the so-called 
(by bridge engineers) Cooper's lines, which are often evidenced in 
the corrosion of bridge members. These are lines of electrolytic 
corrosion in strained parts. The most seriously strained parts, or 
parts that have become hardened in working, are eaten away 
by voltaic action, which goes on at their expense, and the less strained 



358 



STRESS-CORROSION DIAGRAMS, 




^uw -ojm -OJM -ouoii 

GomribiUtj In xMtngj uattM* 



Fio. 50.— Streae-corrosibiuty diagram of cast iron In compressioiL 















" 


r 






40000 














/ 


















y^ 






laooo 


• 

• 








J/ 


/ 










1 






/ 


/ 










MOOD 


e 
1 




J 


/ 














-*J 

• 

e 




t 














16000 


i 
























V 














8000 






\ 


V 




















\ 


V 











•OJ89 



-OJ88 -0.587 -0.586 -OJ85 

CorroilbUltr In acbltrwy units. 



PlO. 67.— Stress-corrosfbility diagram of hard-drawn copper wire. 



HAMBUCHEN'S EXPERIMENTS IN RAPID CORROSION. 359 

parts are less rapidly corroded, thus leaving the appearance of lines 
of corrosion. 

** It is also true that metals which do not change their physical 
characteristics when strained, apparently do not materially change 
in corrosibihty as the result of strain. Thus if lead is stretched, its 
corrosibility does not appreciably increase. The same is true to a 
certain degree of brass and copper. Fig. 54 represents a stress- 
corrosibility diagram for hard-drawn copper. The same is true also, 
to a limited degree, of very soft iron, but as even the softest iron 
does harden sowewhat when strained, its corrosibility is somewhat 
affected by strain." 

An experimental study of the corrosion of iron and steel under 
different conditions has been made by Mr. Carl Hambuchen, B.Sc* 
These experiments were conducted on the lines of those of Mr. Thomas 
Andrews, F.R.S., hereinbefore given, ''on the effect of strain on the 
corrosibility of metal." 

Hambuchen's apparatus and method of testing, and the checks 
and precautions against errors, especially in obtaining the value of 
the electromotive force, were superior to those heretofore employed 
by experimenters, and the results are more in accordance with the 
practical experience of the present day. 

A hindrance to experimental investigation of the corrosion of metal 
is the length of time required to produce measurable results. Ham- 
buchen took advantage of the fact that corrosion being developed 
by an electric current flowing from the ferric body to the electrolyte, 
the rate of corrosion could be greatly increased by causing a current 
generated externally to flow from the metal as an anode, thus causing 
the corrosion to occur under what may be termed exaggerated or 
intense conditions, the metal being corroded as much in a few hours 
as it would be in as many years by exposure to the weather; the 
resultants being practically the same as the effects produced by 
ordinary corrosion. 

The losses in weight from corrosion in different irons, steels, and 
other metals, under strain, from nil to breakage, are tabulated, also 
the loss in weight of metal per ampere hour, and the electromotive 
force developed at various points of the strain. 

Normal solutions of ammonium chloride, ammonium sulphate, 

* Excerpts from Bulletin of the University of Wisconsin, No. 2, E^ngineering 
Series. VoL U, No. 8, July, 1900. Illustrated. 



360 HAMBUCHEN'S EXPERIMENTS IN RAPID CORROSION. 

potassium nitrate, sulphate and chlorides of nitre, were used for the 
electrolyte. The experiments determined that the loss in weight of 
the metal in the different solutions was practically the same, and that 
whether the salts were sulphates, nitrates, or chlorides did not materi- 
ally affect the rate of corrosion. The ampere hours varied from 11 
to 13.6 and the exposures from 19.5 to 24.25 hours. 

Interesting facts were developed showing the variable nature 
of "pitting" in mild steel exposed to different solutions. Figs. 58 
and 61 show a round pitting as the result of the ammonium chloride; 
an elongated pitting results from the ammonium sulphate, and a 
more uniform corrosion from a potassium-nitrate solution. In the 
cast-iron specimens, Fig. 59, the corrosion consisted of a soft carbona- 
ceous material, which generally adhered very firmly to the surface 
of the iron. In case the current density was carried beyond 0.025 
amperes per square inch, this soft material would separate from the 
iron after attaining a certain thickness. The formation of this layer 
must offer some resistance to the flow of the current, and therefore 
protects it to a certain degree. Or, in other words, a given potential 
difference between an iron pipe and a railway track would cause less 
flow if the pipe thus coated were cast iron, than if it were wrought 
iron, which would quickly reveal its weakness by the different char- 
acter of the corroded coating. But with a given amount of current 
flowing from normal cast iron and wrought iron, the corrosion is 
nearly equal in amount in the two. 

It was noticed in the case of the cast-iron anodes that there was 
a liberation of gas during the process of corrosion. That this was 
not due to the flow of the current was shown by interrupting the 
current, the liberation of the gas continuing for some time after. 
The nature of the gas was not determined, nor of what the action 
consisted. 

If the current density was not excessive, the iron did not imdergo 
any material change in appearance, even though subjected to the 
action of the current for a long time. But although the general form 
and outward appearance of the cast iron remained the same, the 
fact that its structure had been materially altered was shown by 
cutting it. The cast iron was found to be softened to a certain 
depth, the material removed having the appearance of fine iron 
filings and graphite, making the loss to the iron equal to 1.17 grains 
per ampere hour for 10.72 square inches of exposed area. This 
coating if allowed to stand until dry became much harder and offered 
some resistance to cutting; but the iron had lost its original strength. 



HAMBUCHEN'S EXPERIMENTS IN RAPID CORROSION. 361 

The effect of the presence of mill-scale on the rapidity of corro- 
sion is shown by Figs. 58-61 and 65 in comparison with the polished 
plates of the same metal. 

Changing the crystalline structure of steel by annealing, hardenmg, 
and burning, causes the amount of corrosion to vary, as is noted in 
the tables on pages 363-364. The amount of corrosion per ampere 
hour of the hardened steel is considerably less than that of the 
annealed or burned steel. 

This is an apparent discrepancy with other data and observa- 
tions in regard to strained metal being the most subject to corro- 
sion. Hardened steel, being necessarily under a highly strained 
condition, should have shown greater corrosibility than the annealed 
or burned specimens. That it did not is owing to the fact that the 
high tension between the particles held them in place until they 
were severally corroded entirely away. In the burned or annealed 
steel the particles, when only partially corroded, were loosened in 
their bond to each other, and cast from the mass before being entirely 
corroded. The pitting of the annealed steel. Fig. 66, the composi- 
tion of which is similar to that of sheet iron. Fig. 72, shows a greater 
corrosibility than the hardened steel, due to the above reason. 

The small percentage of carbon and other impimties in the steel 
would not accoimt for the corrugations in the biuned and hardened 
steel, shown by the figures. It would be unreasonable to suppose 
that the impurities or carbon could be regularly distributed as indi- 
cated by the corrugations; they must be due to lines of strain in the 
metal that corrosion developed. 

The metals that are electro-positive to iron and steel are magne- 
sium, aluminum, zinc, and cadmium; while lead, antimony, tin, copper, 
silver, carbon, manganese, and some of the metallic oxides, are 
electro-negative to ferric bodies. 

Test plates of clean, bright wrought iron, cast iron, and steel 
were drilled and the several holes plugged with one of the above 
metals. The plates were then placed in sand saturated with anmio- 
nium chloride and other corrosive solutions, and after a short expo- 
sure were examined; all the plates in which the electro-positive metal 
plugs were placed were found clean and bright, while the plugs were 
more or less corroded. In the other set of plates the surfaces were 
corroded while the plugs were not affected. In the plates fitted 
with the zinc and other electro-positive metals, the current flowed 
from the solution to the plates and corrosion did not take place. 



362 HAMBUCHEN'S EXPERIMENTS IN RAPID CORROSION. 

In the plates with the lead, carbon, and other electro-negative sub- 
stances, the current flowed from the plates to the solution and the 
plates were corroded. 

The conditions of electrolytic corrosion apply to most of the 
metallic oxides as well as to the metals, and are developed in both: 
1. When two or more conducting substances are in contact with an 
electrol3rte. 2. Whenever there is any difference of electrical poten- 
tial between such bodies. 3. When a suitable connection between the 
conducting substances furnishes a path for the flow of the current. 

All of these conditions are present in the decay of paint coatings 
as well as the corrosion of iron and steel. The electrolyte consists of 
moisture in any form, and may be acidulated, saline, or fresh. 

Iron and steel rea never pure or homogeneous; they contain 
upon their siuHFaces many substances, such as carbon, grapliite, mill- 
scale, and particles of metal and oxides. The body of the metal 
may be formed from scrap iron of different natures, and the heat 
of fusion seldom renders the mass homogeneous. 

In cast iron and steel, in their many processes of manufacture, there 
are many irregular zones of density and purity each of which has its own 
potential. Between all of these differently charged bodies a current of 
electricity is set up, the circuit being completed through the electrol>'te. 

This electric current flowing from the metal to the electrolyte 
v^ill cause corrosion of the metal, which may be general over the 
whole surface of it, or be localize<l in spots, according to its composi- 
tion, which is affected by local disturbances, such as welds, strains, 
annealing, burning, and hardening, or the presence of foreign sub- 
stances. Some peculiar cases of corrosion can be explained by ascer- 
taining if the metal has been subjected to some of these influences, 
that otherwise would bo classed as mysterious. 

Mr. Hambuchen, in order to draw a comparison between electro- 
lytic corrosion and ordinary corrosion, immersed specimens similar 
to those exposed to electrolytic action in a tank containing a normal 
solution of ammonium chloride, and left them undisturbed for four 
months. The results obtained showed that the amount and character 
of corrosion depend upon the quality of the metal, and confirmed the 
conclusion derived from electrolytic corrosion. The time of exposure of 
these specimens was, however, too short to develop any marked pittings 
or other corrosive effects shown so plainly in the electrolytic samples. 

Some of the conclusions given by Mr. Hambuchen as the result 
of his tests are: 



HAMBUCHEN'S EXPERIMENTS IN RAPID CORROSION. 363 



That electrolytic corrosion produced by the flow of a current of 
moderate density from an external source, produces results on the 
metal which are similar to those produced by corrosion under ordinary 
conditions. 

In many if not all cases the character and rapidity of ordinary 
corrosion of iron and steel depend upon their physical and chemical 
properties, and the galvanic action due to differences in potent al 
between different parts of the metal. 

The application of stress to metals causes an increase in chemical 
activity, this increase being especially marked after the elastic lunit 
is reached. 

It is possible to plot a curve showing the relation of electromotive 
force to strain, which is similar to that of stress to strain. 

There is a definite relation between the electrical potential of 
any metal toward an electrolyte and the amount of energy stored up 
in the metal through the application of stress. It is evident that the 
protection of ferric structures from corrosion requires their removal 
from electrolytic influences. 

The several specimens subjected to different conditions of corro- 
sion were all taken from the same bar or sheet to facilitate comparison. 

The follo\^'ing tables are means of a few of the separate results 
given by Mr. Hambuchen: 

Table 9ho\^'ing the loss in weight of iron and steel used as anodes, immersed in 
a solution of ammonium chloride and exposed to the action of an electric current 
of varying densities and time. 



Material and Condition of Surface. 



Annealed steel, polished 

" " with scale 

Hardened steel, polished 

" " with scale 

Steel bunied, not hardened or pol- 
ished .• 

Steel burned, not hardened with scale 

Steel burned, hardened, and polished 

Cast iron, polished 

" " scale partly removed .... 

Sheet iron, polished 

" " scale partly removed. . . 



Area in 
Square 
Inches. 



10. 

10.523 
10. 
10.48 

10. 

10.3 

10.555 

10.073 

10.693 

10.373 

10.263 



Total Lofls 
in Weight. 
Grama. 



15.9 
15.33 
14.633 
14.407 

15.666 
15.515 
12.875 
12.273 
9.996 
13.87 
14,63 



Weight 

LoBt per 

Ampere 

Hour. 

Grams. 



1.1683 
1.1163 
1.077 
1.059 

1 . 1526 

1 . 1575 

1 . 1475 

1.0983 

0.7506 

1.184 

1.084 



Ampere 
Hours. 



13.6 
13.3 
13.6 
13.3 

13.6 
13.3 
11.2 
11.2 
13.3 
11.2 
13.3 



Ex- 

g)sure 
ours. 



23} 
19} 
23i 
19.5 

23} 
19} 
24} 
24} 
19i 
24} 
19i 



It will be noticed that the amount of corrosion for all of the specimens is 
greater per ampere hour than the theoretical amoimt given by Faraday's law 
(1.0448 grains per square foot of surface per ampere hour), cast iron with the 
scale partly removed being an exception. 



HAMBUCHEN'S FIGURES OF CORROSION. 



HMd and CoDditioD. 


s. 


Total La» 


sSk. 




10.437 
10.30 
10.556 
13.596 
14.16 
14.117 
10.37 
10.603 
10.336 
10.57 
10.255 


0.733 
1.473 
1.793 

1.083 
1.663 
1.663 
0.733 
1.08 
2.00 
1.063 
3.70 




























" " ' Bcale partly removed. 


0.102 






Steel with scale, burned uot hardened 


0.3613 



Fw. fiS, — Mild steel (ammonium-ohloride solution). Magnified 2J diameters. 



Fig. 59.— Cast iron (polished). Magnified 2J diameters. 



HAMBUCHBN'S FIGURES OF CORROSION. 



a (with scale). Magnified 2} diamehTB. 



Fig. 61.— Mild steel (ammonium- chloride solution). Magnified 2i diametei& 



Fio. 62. — Burned and hardened steel. Magnified 2) diameters. 



HAMBUCHEN'S FIGURES OF C0BBO3I0N. 



Tio. 63. — liiimed steel not hardened (with scale). Magnified 2J diametra. 



Fig. 64. — Annealed steel (polished). Magnified 2i diameters. 



Fia. 65. — .Annealed steel (with scale). Magnified 21 diameters. 



HAMBUCHEN'S FIGURES OF CORROSION. 



Fia. 66. — Steel burned but not hardened (polished). Magnified 2} di&inet«rs. 



Fia. 67. — Hardened steel (polished). Magnified 2} diametera 



Fio. 6S.— Hardened steel (with scale). Magnified 2| diameters. 



HAMBVCUEN'S FIGURES OF CORROSION. 



FiQ- 69.— Steel bumed and hardened (polished). Magnified 2J diameters. 



Ite. 70. — Sheet iron (polished). Magnified 2} diameters; 



Fw. 71.— Sheet iron. Magnified 21 diametera. 



HAMBUCUEN'S FIGURES OP CORItOSlON. 



Fio. 72.— Sheet iron (nith scale). MaRoified 2j diameters 



CHAPTER XXXIV. 

ELECTROLYSIS OP UNDERGROUND METAL. 

Prop. Edwin J. Houston defines electrolysis as chemical 
decompositioii effected by means of an electric current; that is, 
the source of electrolytic corrosioa lies in the release of an atom of 
oxygen from any moisture and its instantaneous combination with 
any iron it can seize upon." 

Prof. D. C. Jackson (University of Wisconsin) defines its action: 
"In an electrolytic cell with iron electrodes, having any salt or salts 
of alkaUne metals or earths in solution in the electrolyte, the salts are 
decom[X)sed, their acid radicals attacking the anode, forming an 
iron salt. 

A proof of this theory is found in the storage battery, in the charg- 
ing of which a large current is discharged from a lead plate into an 
electrolyte consisting of dilute sulphuric acid. In a storage batter^' 
the oxidation of both plates takes place but the red oxide of lead 
formed is very different in character from the corrosive effects of an 
electric current on an underground lead-pipe." 

* "There is nothing mysterious about the corrosion of metals b}*^ 
electrical currents. Its action is precisely similar to that employed 
by electro-platers in their art. Two plates of metal placed in any 
material, whether damp earth or the solution vat of an electro- 
plating apparatus, a current of electrical energy will be instituted 
from one plate to another and the plate or object from which the* 
current flows will be corroded. The current will make its own selec- 
tion as to its course and the body to be attacked. In the case of 
metals buried in earth or water, the electrolytic action is out of sight 
and probably out of mind, but none the less present and uncontrolla- 
ble; while in the plating bath it is in sight and controllable more 
or less, at the will of the attendant. In the case of high voltage and 
large ampere currents returning to their source of generation, it is 
one of the laws of electricity that where a current has two paths to 

* "Electrolyeifl," Engineering Record, August 21, 1899. 

370 



ELECTROLYSIS IN FERRIC BUILDINGS. 371 

reach this point (and all electrical currents have two paths), the 
current will divide and return to its source in the direct ratio of their 
conducting capacity, whatever these conductors may be. Even 
with large and well-bonded metal return conductors biuied in the earth 
or in conduits, some of the current will invariably pass by way of the 
earth and reach any outlying metaUic bodies, The low voltage of .001 
to .01 and the amount of amperes will determine the rate of corrosive 
effect in all metal in their course." 

The advent of the steel building almost simultaneous with the 
introduction of the dynamo, has added not another form of corrosion, 
but a new field for its development and a new danger. 

The principal part of the metal in steel frame structures is so 
embedded in masonry as to render inspection of its condition almost 
impossible. The pipe systems are more accessible, but are never- 
theless at all times a ready prey for electrolysis. 

Three hundred horse-power of electrical energy are not uncom- 
mon installations for Ught and power in one building. Whether led 
in from the outside, or generated in the building does not change 
the effect, which is to disturb the normal electrical conditions of all 
metals in the immediate neighborhood and in many cases those far 
distant. 

In the return of this energy from its work to its generating source, 
if the pathway is not made perfectly free by the use of a conductor of 
adequate size, or if it be of such a length as to render a shorter and 
better circuit through other objects possible, then the current will 
jump the line wholly or in part. On the new route, wherever it leaves 
the metal, another jump will occur, and the metal will be corroded at 
that place, and not where the energy entered. There is always 
moisture enough in any building to afford adequate oxygen for the 
corrosion. 

In the steel frame work of the building, the electrolysis will be at 
the foot of the columns nearest to the least resisting pathway of the 
current, generally at a point impossible to locate or inspect, and 
with moisture in excess to make a good cross-cut and dangerous 
circuit. 

In the iron pipes there usually will be a jump at every joint, if it is 
made by the hemp gasket and lead, or by a hydraulic cement method. 
The screwed ends of wrought-iron gas and water-pipes also are affected 
by the difference in resistance between the two natures of the same 
metal, and sheet iron used for shims between columns and girders 



372 ELECTROLYSIS IN FERRIC BUILDINGS. 

have been found to be the cause of a jump corrosion, particularly 
if aided by the presence of damp dirt or other substances. 

A voltage of 118 maintains an arc of one inch in an electric arc 
light, and requires 1 H.P. of electric energy, and fourteen six- 
teen-candle-power incandescent lamps represent the same amount 
of energy. It is not infrequent to find an arc light or twenty or 
more incandescent lights, or an equivalent energy from small motora 
coimected to the pipe systems or the steel frame of the building. 
The effect of these strong currents is to set up a corrosion in the steel 
at some point where the current is interrupted on its return to the 
dynamo. 

Under some one of their many developments, induced currents 
are strong enough to corrode metal, even if the main current or return 
current wires are adequate for their duty. Less than .005 volt estab- 
lishes electrolysis, the amount of which is in proportion to the amperes 
present and not to the voltage. 

Protection from the effects of the jump of the current, particularly 
in the lower parts of a steel frame, is rendered more uncertain, by 
reason of the disturbing action of electric currents from adjoining 
sources of generation. These currents finding their return to their 
source of generation obstructed from any cause, form a short or 
easier circuit that often lies through another dynamo's sphere of 
action. They invade its field and disarrange a return current system 
that at first might have been adequate for its duty, but is not able 
to withstand currents from its neighbor of subsequent installation. 

Twenty or more of these electrical installations of different degrees 
of voltage, ampere, and work, are often placed within a compara- 
tively small area. Many of these have been found to have a return 
wire system of inadequate size or of faulty insulation, and all of them 
subject to a wide range of fluctuation in energy due to the varying 
character of their separate work. With these influences at work at 
nearly all hours, it may be confidently expected that the near future 
will reveal some large and dangerous examples of corrosion in steel- 
framed buildings, from sources that thus far have received only a 
cursory consideration and no prevention. 

Many instances are on record showing the erratic action and dan- 
gerous character of either stray or direct electrical currents. The 
United States Astronomical Observatory at Washington, D. C, 
though eligibly situated and free from the disturbing influences of a 
laige city, had its magnetic observ^ation department rendered useless 



ELECTROLYSIS IN FERRIC BUILDINGS. 373 

by the stray electrical currents from a trolley line of street railway 
some three-fourths of a mile distant. This branch of science so 
closely associated with the daily needs and welfare of mankind 
was paralyzed by the culpable indifference of a corporation to the 
requirements of science. It took an Act of Congress, carrying the 
imposition of a heavy fine to stop the nuisance. The Act not only 
prohibited the use of any underground water- or gas-pipes as means 
for returning the trolley-line currents to the power house, but also, 
forbade the connection of either pole of a railway dynamo in any 
direct manner with the earth. 

Paints furnish neither renaedy nor protection from electrolysis. 
Paints under catchy nanies are extensively advertised as being elec- 
trically inert, or insulating in charcter. Such names and state- 
ments are misleading and unreliable. No paint whose pigment is 
an oxide and again reducible by heat to a metal, is non-electric or 
passive to electrical influences in any degree beyond that due to 
the difference between the oxide and its metal, generally about 50 per 
cent, but is never nil. 

Lampblack and graphitic carbon are the only pigments that are 
partially non-electric, and even with the use of these in a paint, the 
coating as an insulating substance is governed by the vehicle. The 
vehicles containing the resins, fossil gums and refined bitumen and 
combined by heat into a varnish, are the best for non-electric paints. 
It is quite unusual to find them in use on account of their cost, while 
the cheaper grades of resins and resin-oils used in the vehicle are 
only insulating up to a certain percentage, when they become con- 
ductors. 

All of the vitreous class of pigments, such as slags, hard-burnt 
brick, tiles and slate, are conductors in their pulverized form, and 
usually act in a paint as the negative electrode to concentrate the 
electrical enei^ upon the covered ferric body. The thin coating of 
the vehicle, jtj^o-j to j^ inch in thickness, is not resistant enough 
to but partially insulate the pigment, however effective the vehicle 
may be in mass or in heavier coatings that could not be applied cold 
with a brush. 

Electrolysis inaugurated beneath such coatings, generally throws 
them off, or they act as a mask to conceal its ravages ; while inferior 
coatings are rendered hard or porous from the decay of the vehicle or 
pigment. 



374 ELECTROLYSIS OF THE PEOBIA STAND-PIPE. 

* The following cuts and descriptions illustrate an instance of the 
effect of corrosion induced by stray electrical currents that caused 
the destruction of a water-works stand-pipe at Peoria, 111., March 30, 
1900,,with the loss of life of two persons and the injury of fourteen 
othere. The stand-pipe was 60 feet away from the other stand-pipes, 
and more than a mile away from the power station. Fig. 73 repre- 
sents a sample of a steel sheet from the stand-pipe, showing the pit/- 
ting in the sheet around the edges of the rivet-heads. The examina- 



Fia. 73. — Htting of steel stand-pipe sheet around the rivets. 

tion of the wreck of the stand-pipe showed that the whole inner sur^ 
face of the vertical shell appeared to be thickly covered with blbters 
resembling in outward appearance the tubercules sometimes found 
inside of old cast-iron mains. 

A similar stand-pipe on the East Bluff was drained, and was found 
to be similarly pitted. Tliis blistered covering, which was almost 
as thin as paper, was composed entirely of oxide of iron, and on 
brushing it away the black paint with which the stand-pipe had been 
originally coated was found beneath it. The paint was oftentimes 
almost unbroken, or. at least, very slightly cracked. When the 

• Excerpts from "Electrolysis of UnderRround Metallic Structures." A paper 
read by Mr. Damey H. Maury, Chief Engineer of the Peoria (I1L> Water Works, 
before the American Water Works Association, May, 1900. En^iurering News, 
June, 1900; atao, American Gas Light Journal, July 30, 1900. 



ELECTROLYSIS OF THE PEORIA STAND-PIPE, 375 

paint was brushed off the pit would be disclosed, considerably 
smaller in area than the surface covered by the blister. The surface 
of the metal in the pit was perfectly bright and clean and its fibre 
was clearly discernible. Many of these pits were more than J inch 
in depth. They were slightly more numerous in the West Bluff stand- 
pipe than in the East Bluflf stand-pipe, and were in both generally 
larger and deeper on the lower courses of the vertical shell. 

The electrical examination relative to the stand-pipes was conducted 
mainly at the East Bluff stand-pipe, which was still in service. 
A flow of a part of the current from the railway-line was clearly traced 
through the earth to the anchoi>bolts which held the stand-pipe to its 
foundation, as shown in Fig. 74, up these bolts and into the steel of 
the shell, and through the shell and from its inner surface to the 
projecting section of the 16-inch flanged cast-iron pipe which served 
as both inlet and outlet, and which connected the stand-pipe to the 
water-mains. The current was then traced along this pipe and 
along the mains to the power station. The deflections of the 
volt-m.etre needle were clearly traced to the railway current, being 
especially influenced by the cars on the line beyond the stand-pipe, 
and when the cars stopped running at night, the movement of the 
needle ceased. Where the current left the inner surface of the shell 
to pass through the water to the inlet pipe it made the pits already 

described. 

Fig. 78 shows the interior surfaces of three sections of this inlet 
pipe, marked A, 5, and C, respectively, the positions occupied by 
these sections originally being shown by the lettters A, B, and C in 
Fig. 74. An examination showed that strongly marked and numer- 
ous pits were inside the sections A and 5, while the inner surface of 
the section C was practically as smooth and perfect as though new. 
When the condition of the inside of these three sections of pipe was 
first noted, it was hard to understand why A and B should be pitted, 
while C was unaffected. A closer examination, however, showed 
that in the flanged joints between the bottom sheet of the stand-pipe 
and A and 5, respectively, corrugated copper gaskets were used, 
while the pipe B was separated from the pipe C by a thick rubber 
gasket; and that under the nuts and heads of the bolts holding the 
flanges together, there were grummets or wrappings of cotton wick 
soaked in tallow. 

The result of this arrangement was, that the current which entered 
A after passing through the water from the inner side of the shell 



376 ELECTROLYSIS OF THE PEORIA STAND-PIPE 

of the atand-pipe, and which was trying to return along the inlet 
pipe and water-mains to the power station, encountered, at the joint 



Fia. 74. — Partial section of Peoria stand-pipe, ahowing course of electrical 



between B and C, the rubber gasket and the grummets. The effect 
of the gasket and grummets was to practically insulate the section C 



Fio. 75. — Interior v 



from the sections A and B. and as none of these pipes were in contact 
with the ground, the current was compelled to leave the pipes A 



ELECTROLYSIS OF THE PEORIA STAND-PIPE. 377 

and B and travel through the water or along the slimy coating of 
oxide on the inside of the pipes around the joint between B and C, 
in order to continue on its journey. As the current was not leaving 
C, this pipe was not injured, but the current, in leaving the inner 
surfaces of A and B, did pit them, as shown in the photograph. 

The experiments conducted since the destruction of this stand-pipe 
have determined that no manner of packing the joints in an under- 
ground cast-iron water- or gas-pipe line affects, only in a small degree, 
the difference in potential between the two ends of the connected 
pipes. Whether the joints were well or poorly calked, the pipes 
empty and dry, or full of water, clear or muddy, with scale on the 
pipe or clean surfaced, the drop of potential around the joint only 
varied from 0.0145 to 0.008 volt, and the general average resistance 
of the joints was about 96 per cent of the resistance of the whole line 
of pipe, and the resistance of the joints increased with age. 
Pitting was always observed where the shunt of the electric current 



I water-main. (Current 

left the metal to flow around the joint, and this corrosive action was 
as marked upon the inside surfaces of the pipes as upon the external 
surfaces; but from the conditions could not be observed. 
Wrought-iron pipes joined by the usual screw-thread and thimble 
connections were almost as universally attacked by electrolytic 
action at the joints of the cast-iron pipes. The difference in potential 
between the pipe and the socket, from their different arrangement 
of metallic fibres, resulted in the faster corroding of the pipe ends 
than the screwed socket or thimble. 



378 ELECTROLYSIS OF THE PEORIA STAND-PIPE. 

The joints in the cast-iron pipes, whether coated or not with the 
usual coal-tar preparations put on at the pipe-foundry, had Uttle or 
no effect to insulate or protect the pipe from electrolytic corrosion, 
which generally showed in the form of blisters or like the usual tuber- 



cules formed on water-pipes by the action of water. These tuber- 
cules when broken, and the material under them analyzed, showed 
22.3 to 23 per cent graphitic carbon and 47 to 47.7 per cent iron. 

The soil in which the pipes were embedded had received and 
was impregnated with the metal (iron or lead) corroded by the electric 
current, even to some distance from the pipe. In sections of 
pipe from 600 to 2000 feet long, there were several zones and centres 
of greater action than on the pipe in general, but they all centred 
toward the nearest generating dynamo, and were p\idently localized 
at these points, from the greater amount of electrical enei^y cast off 
at the meeting and passing points and switching of the cars; con- 
tiguous lines of cars and currents from other sources all contributing 
in the most erratic manner to the corrosive result. 

Mr. William Work, in a communication to the Institution of 



ELECTROLYSIS LN WATER PIPES. 379 

Civil Engineers, 1901, reports the rapid corroding of wrought-iron 
service-pipes for water and gas, laid in a light, sandy soil. The 
water service-pipes corroded in seven years, so as to need renewal, and 
the gas service had almost completely disappeared at the end of 
ten years. 

Analysis of the soil disclosed the presence of common salt, mag- 
nesium chloride, iron, alumina, silica, lime, and posphates. The town 
where the pipes were laid had no system of sewers, and during the 
summer season the streets were daily watered, and as the streets were 
level the water was quickly absorbed. The subsoil in which the pipes 
were laid was porous and alternated from dry to damp. 

Carbonic acid was generated from the chemical action of the 
soil and attacked the pipes as stated. The trouble was confined to 
the wrought-iron service-pipes; the cast-iron pipe-mains to which 
the service pipes were connected were not affected in any noticeable 
degree. Similar pipe-services laid in neighboring towns where the 
soil was of a decomposed granite nature were comparatively unin- 
jured after a period of twenty- even years. 

Salt or lime in any soil in which pipes are laid necessarily prove 
active agents to promote corrosion, as they are hygroscopic in nature; 
and if alkaline substances are also present in the form of ashes and 
coal cinders, as they nearly always are in the soil of towns, the life of 
all wrought-iron work buried in it will be very short, even with 
the usual crude coal-tar coatings. Clay puddle around such pipes 
proves a good protection, as pipes so protected have been found 
practically uncorroded after forty years. 

Mr. Chas. W. Rowe, Secretary of the Dayton, Ohio, Water De- 
partment, reports that the cast-iron water-mains laid in 1891 
were found in 1900 so greatly affected by e ectrolysis as to endanger 
the water-supply of the whole city. Voltages of 4.5 were found in 
many parts of the pipe system. 

The Annual Report of the Water Department shows that in 1899 
579 feet of 6-inch pipe and 26 feet of 4-inch pipe were abandoned on 
account of electrolvsis. 

Mr. Rowe reports that in 1898 over 46,000 feet of the water- 
pipes from 4 to 16 inches in diameter were so seriously corroded by 
electrolysis from the trolley-line currents, that in some cases over 
one-half of their strength was gone. A 6-inch pipe became useless in 
five years. The pipes became coated with a graphite-like sub- 
stance tV or more inch thick, the pebbles and stones in the groimd 



380 ELECTROLYSIS IN WATER PIPES. 

near the pipes being plated with the same substance. About J J of 
the electric current used to drive the trolley cars was found to pass 
through the pipes and only ^ passed by the street rails on its return 
to the dynamo. 

Fig. 78 shows the electrolysis of a 4-inch cast-iron water-pipe at 
Reading, Pa. 



Fig. 78. 

No. 1. Laid 22 years or about 18 years before the advent of the 
trolley lines at the point where the pipe was laid. 

No. 2. I^id thirteen months. 

No. 3. Laid 22 years. Burst when uncovered for examination. 

Fig. 79 shows a 16-inch cast-iron suction pipe; also a water main 
from Reading, Pa. The pipes are samples of those laid 300-800 
and 1000 feet distant from the street trolley lines, and all were over 
two miles distant from the electric power station. The effect of 
the resistance of the joint packing is seen in the corrosion of the 
ring on the end of the pipes; also in the general corrosion at the 
en<ls where inclosed in the bell. 

Fence nails were driven into the ends as easily as into a wooden 
post. 

Flakes of corroded metal 3"xl" in size and ^inch in thickness 
were of frequent occurrence in the IVailing pipes. 



ELECTROLYSIS IN WATER PIPES. 381 

"Electrolysis of Underground Water-Pipes " by Mr. F. C. 
Kelsey, Chief Engineer, of Salt Lake City, Utah, is a condensed report 
of the reports from the chief engineers of seven cities in the United 
States, of the presence of electrolysis in their water-supply systems. 



Fig. 79. — 16-inch caBt-iron suction pipes (page 380). 

Briefly stated, electrical currents of low potential, .001 volt, 
were found and were enough to establish electrolysis that under 
favorable conditions for the pipe might not have become serious in a 
limited period, but in the case of soils carrv'ing salts and acids of 
decomposing materials the corrosion from electrolysis was materi- 
ally accelerated and increased rapidly as the voltage rose to .01 volt. 
Currents of one-half volt destroyed a telephone cable in a few months. 
A minute quantity of soluble salt in the soil was enough to start the 
current and establish the electrolytir point or points in the metals, 
the corrosion of which continued as long as the current flowed. In 
all cases where the water-pipes or the metal-work of the building was 
positive to the earth or any surrounding object, the electric current 
flowed in that direction and electrolysis was established in the posi- 
tive element of the couple, wherever it was situated. Tt was noticed 
in many cases that the lead service pines and the lead-packing in 
the water-mains were corroded to such an extent that excessive 



382 ELECTROLYSIS L\ WATER PIPES. 

leakage was the result, and the pipe-joints were no lonj^er able 
to resist the pressure of the water wnen it was over 20 pounds. This 
corrosive action upon lead is of ioterest, aa metallic lead has been 
generally considered electro-passive, aad is used for the outer insulat- 
ing covering for the cables in all underground conduits for electric 
lighting and power. 

The Report of the Board of Commissioners of Klectrical Subways, 
Brooklyn, N. V., 1894, states that nearly 300 miles of lead-coated 
telephone cables were rendered useless by electrolysis from the trolie^- 
currents in that year. Many cases of corrosion were found where 
the lead cable was incased in pitch and other insulating compounds. 

Mr. A, A. Knudson, E.E., reporting the electrolysis of a 48-incIi 
diameter water-.nain in the city of Cambridge, Mass., found vol- 
tages of 25, and amperages of 30-50-80 and 90 at many points of the 
water-supply system. 

Fig. 80, Electrolysis of a 6-mch cast-iron water-pipe at Provi- 
dence, R. I. The pipe was i-inch thick when laid and had been in 
service seven years. 



Fio. 80. — Electrolysia of a 6-inch cast-iron pipe at Providence, R. I. 

Fig. 81 shows the electrolysis in one end of a steel truss for a 
trolley railway bridge at Providence, R. I. Both ends of both trusses 
were similarly corroded near the ground line. 

The special committee of the American Water Works Associa- 
tion, to whom the subject of "Electrolysis of Underground Wat«r- 
pipes" was referred, reported at the Richmond, Va., meeting of 
1900, the result of their investigations; "That with the best bonded 



ELECTROLYSIS IN WATER PIPES. 3S3 

<X)imeGtioii of the rails, includlDg even the welded joint, it was 
impossible to secure the return of all of the current from a single 
trolley railway-line to its source of generation. Some of the current, 
under the law of divided currents, would invariably leave the rails 
and seek another source of return through neai^by met&l. The 
smallest measure of difference in potential between two metallic 
bodies was sufficient to produce and maintain electrolysis in one 
of them." 



Fio. 81. — ElectrolyMs of & eteel bridge truss. 

Dr. Leybold's paper, "Electrolysis of Gas-pipes,"* states: "The 
pipes when Itud were protected with canvas soaked in boiled coal- 
gas tar, and the destruction of the pipes was more rapid than where 
they were laid without the canvas coating. New pipes laid with 
canvas and tar coatings to replace the old ones were perforated into 
holes in seven to eight months." 

The annual report of Mr. Wm. Jackson, City Engineer of Boston, 
Mass., states that -j^ of a volt was sufficient to cause electrolysis, 
and some of the most serious cases of electrolytic action in the city 
water-mains were where only 1.5-voIt pressure existed between the 
ground and the pipes. 

* American Gas Light Journal, Sepleiiiber 30, 1901, p. 526. 



iKS* ELECTROLYSIS OF OAS PIPES. 

Mr. L. Holman, Water Commissioner tor the city of St. Louis, 
Mo., calls attention to the corrosion of 48-inch diameter water-mains 
in that city, that have been eaten away in many places for one-half 
inch, and couJd be cut as easily as plumbago. The danger arising 
from the bursting of such a pipe is apparent. 

The practical effect in the corrosion of undei^round gas-mains 
and wrought-iron service-pipes, principally those of small diameter, 
is noted in the Official Pleports of the Gas Bureau of the city of 
Philadelphia, where the loss from leakage in the form of unaccounted- 



Fio. 82— Exterior of a pipe injured by plectrolysis. Springfield. 111. 
for gas, for a period of ten years, was $.5,750,000, and for the years 
1891 to 1895 averaged two millions of cubic feet per day. 

The Brooklyn Union Gas Company (Brooklyn, N. V.) has about 
760 miles of gas-mains of all sizes, and 280 miles of wrought-Jron gas 
service- pi pes. The latter and their fittings are found to be badly 
corroded wherever uncovered for examination. Electrolysis from 
stray electrical currents is manifest in many cases. Thirty-eight 
service-pipes in one street block were completely destroyed in three 
years. The cast-iron mains are reported to be generally in a good 
condition so far as electrolytic action is concerned, except in a few 
cases in what is called "the dangerous district;" that is, in the 
vicinity of the electric-station power-houses. 

The loss of gas from all of the underground systems in this city 
in 1899 amounted to 13 per cent of the total yearly output (4,500,- 
000,000 cubic feet), or a loss of 585,000,000 cubic feet. 

Other cities in the United States show similar conditions in their 
gas systems. The ordinary corrosion of imderground metal has 
been materially increased since the advent of electric street railwaj-s, 
however thoroughly the rails are bonded and used for the return 
current. 



ELECTROLYSIS OF WATMK AND OAS PIPES. 38S 

Lengths of pipe-mains over three miles long are reported to have 
corroded to such an extent that the whole pipe-hne had deteriorated 
50 per cent in four years. The voltage in this line was from 2 to 9 
positive. At a voltage averaging 4.5, a ft-inch pipe became useless 
in five years. 



Water-mains that were laid and tested to withstand a pressure of 
over 300 pounds to the square inch, at the end of four years leaked at 
almost every joint at 150 pounds' pressure. The voltage was 4.5, 
and the lead-joints were ba<lly corroded. 

Electrolysis of water-pipes at Kansas City, Mo.,* Mr. G. B. Wing, 
Superintendent of the Metropolitan Water Works, reports that speci- 
mens of the soil at a distance of two inches from some of the corroded 
pipe showed 4.67 per cent of iron, and at a distance of one foot, 2.65 
per cent. 

The eorrosion of wrought-iron and lead pipes was more rapid than 
that of cast-imn pipes; the amount of corrosion in all cases depended 
upon the amperage of the current. 

The electrical resistance of the ordinary lead and oakum-packed 
joints in the pipes was found to range from 110 to 200 times that 
of the pipe itself. 

Variations in the resistance at the joints ranged from 0.0264 to 

•Engineering Record, Vol XL, No, 11, August, 1899, p. 239. 



386 ELECTROLYSIS OF WATER PIPES AND STREET RAILS 

0.032'2 ohni, and in a section of 25 lengths of 'l-iiich pipe, averaged 
0.ii ohm. 



Fig. 84.^Effect of electrolyws on 6-inch cast^ron pipra, Kansas City, Mo. 

"Wandering Electricity in New York City."* A report by 
Mr. A. A. Knudson. E.E., of an electrical survey of the section of 
New York City at Third Avenue and 135th Street, where a trolley- 
line had a terminus in front of an elevated-railway station. There 
was a difference of 2 volts, rising to 10 volts at times, between the 
trolley tracks and the nearest rwlway column, and a difference be- 
tween the trolley tracks and the nearest gas-main of 5 volts. 

In removing the rails of the trolley-line the effects of the passage 
of the current from them to the gas-main and elevated-railway struc- 
ture were shown by the corrosion of the 70-pound rails. The base 
of the rails, originally 4 inches wide, was corroded away to 2? inches 
wide near the en<Is, the edges of the flanges being corroded to knife 
edges for several feet back from the ends. 

Fig. 85 shows a section of the rails at one end. 




Fifl. 85. — Electrolyais of a street-railway steel te«-rail. 

The wToughtr-iron gauge-ties originally li"X|" in section were 
corroded completely away in the centre, and but one was found near 
the station that was imaffected from end to end. 



• Entpneering Record, VoL XXXVIII. No. 23, November 5, 1898, p. 500. 



ELECTROLYSIS OF STREET RAILS. 387 

In the current at the terminal that passed to the water-mains 
and to the elevated-railway structure, there was a difference of from 
2 to 2.6 volts. At one-fourth of a mile away the current all passed 
to the elevated-railway structure, ran a fourth of a mile, then 
returned to the water-pipes and changed again to the railway struc- 
ture in about half a mile. 

The same conditions were found to prevail on an opposite section 
of the elevated-railway structure, extending for about a mile in the 
opposite direction from the trolley terminal. 

The difference in potential between the elevated-railway columns 
and the street-trolley rails and water-mains ranged from 1.30 to 1.50 
of a volt, and indicated that the current came from an electric light- 
ing station. It was also shown by the tests that a trolley-line using 
the rails, water and gas-mains for its return service, can spread the 
corrosive influences for a mile in either direction through the various 
subway conduits, pipes, and elevated-railway structures; also that 
the conductivity of a 50-pound street or tee-rail is about equal to 
a copper rod 1 inch in diameter, or five No. 000 B. and S. copper wires. 

Tests applied to the Brooklyn Bridge suspension cables showed 
that generally there were 3 volts positive to the rails of the trolley 
railway on the structure. The effect of these currents upon the 
anchorages of the bridge led to a number of tests of the upper ends of 
the anchorage metal. The tests are believed by the bridge engineers 
to show ''that no damage such as might be expected from corrosion 
of underground metal has thus far taken place.'' 

A wise distinction between corrosion and electrolytic action. 
Had the question of the corrosion of the rails in the street tracks been 
put to the trolley-railway engineer corps, they would probably have 
been positive that no such corrosion was present or possible; in fact, 
they were indifferent to or ignorant of the corrosion until the exam- 
ination by Mr. Knudson. 

It required the public evidence of a half-dozen of broken suspen- 
sion rods and panels of sunken railway tracks to convince the Brook- 
lyn Bridge engineers that a serious case of neglect and corrosion 
existed in the structure, and had progressed far enough to be dan- 
gerous to it, ere a few long-neglected repairs were made. 

That that neglect does not include the anchorage metal is by no 
means certain. There is absolutely no plan in the many suspension 
bridges erected that affords any practical means to ascertain the 



888 ELECTROLYSIS OF SUSPENSION-BRIDGE CABLES. 

state of the lower members of the anchorage system, or of arresting 
corrosion or electrolysis in them if found. 

Iron and steel bodies exposed to conditions similar to bridge 
anchorage metal have been found badly corroded within a few years 
after being placed in position, and there is no reason to infer that 
any bridge anchorage will be an exception. 

The rate of corrosion from natural causes has been fairly deter- 
mined. In anchorage work this will be increased by any electrical 
currents that may reach them, and it is inevitable that they do reach 
them, and no means of preventing it now exists. 

The decay of metal by electrolysis has been approximately ascer- 
tained. The escape of the voltages and amperes used in street-rail- 
way service is twenty times that necessary to induce ferric corrosion 
and often more than twice as much as is necessary to decompose 
water in mass. 

A current of 0.3 ampere is sufficient to corrode a lead covering 
to a cable or the lead in a pipe-joint. Electrical engineers report 
cases where the lead covering of cables has been destroyed in six 
weeks after laying. 

A potential of jjf^ of a volt is all that is required to induce ferric 
corrosion two miles from the dynamo. 

A difference in voltage of 20 volts has been found between 
the two ends of the Brooklyn Bridge cables, and the difference in 
voltage ranges from 0.75 to 3 volts at all hours and at all times in the 
day whenever tested, and is always found electro-positive to the 
ground. 

So long as electricity obeys the known laws pertaining to its genera- 
tion and transmission, it will select the line of least resistance, though 
it may not be the shortest in returning a major part of the current 
to its individual source of generation. It will also divide en route, 
pick up other electric currents in the most erratic manner, and deposit 
them in unexpected places, generally inaccessible for observation 
or repairs. 

The large amounts of voltage and amperes used in railway-motor 
systems render stray electric currents more certain and electroh-sis 
more constant, even if a "shunt" of the current from any adjoining 
bridge cable or structure were possible. At the present state of the 
electrical art this "cut-off" is practically not feasible. 

The river that separates the two anchorages compels the bridge 
cables to act as conductors for the electric currents present at all 



ELECTROLYSIS OF SUSPENSION-BRIDGE ANCHORAGES. 389 

times in the earth and air and never, or but momentarily, of the same 
potential. 

Hundreds of electric installations of a great diversity of power 
surround these bridges and provide a cause of danger that at present, 
if known or suspected, has no remedy or safeguard. 

The future results of electrolysis on all suspension bridges may 
as well be recognized now, rather than be left till the inevitable catas- 
trophe befalls. 

The corrosion of ferric bodies, not aided by electrolysis, is known 
to be progressive, being nearly 50 per cent more the second year 
than the first, and so on for each succeeding year. 

During the construction of the Britannia Bridge over the Menia 
Straits, some rejected plates jV and f inch thick, were left unpro- 
tected and exposed to the spray and wash of the sea. In two years 
they had corroded so that they could be swept away with a broom. 

A few pieces of ironwork embedded in mortar or walled in some 
ancient building, or an old water-gate here and there, in some very 
favorable situation, may have remained uncorroded, but there is 
little imquestionable proof that iron or steel in the form adopted 
for bridges or structural frame-work will last more than two him- 
dred years. 

In the Niagara Falls and the Alleghany River suspension bridges, 
after about twenty-five years of duty, an inspection showed that 
some of the wires in the outer strands of the cables outside of the 
anchorages were corroded through, but the second and interior wires 
were sound. The reason assigned for the corrosion of the outer strand 
wires was that the "creep" of the individual wires under the varying 
strains due to the load and constant changes in temperature had 
worn away the boiled linseed-oil and other coatings applied when 
the wires were strung and allowed atmospheric moisture to reach 
them. 

It is now proposed to abandon the boiled oil or paint coatings of 
the cables and to use a mixture of vaseline and pliunbago. When 
the wires are strung and ready to bunch into strands and cables, all 
of the interstices are to be filled as far as possible with this stiff un- 
drying mixture, that is to act as a lubricant for the inevitable creep 
of the wires, also as a protection from corrosion. 

In all wire-wrapped cables there is an appreciable space between 
the wrapping and the cables, caused by the drying and shrinkage 
of the boiled oil coatings applied during their construction. The 



390 CORROSION IN SUSPENSION BRIDGES. 

daily changes in temperature of the cable wrappings are greater than 
the mass of wires they cover, hence with the changes due to the 
extremes of summer heat and winter cold, they necessarily prove an 
element of weakness in providing a foundation for the paint coat- 
ings that are supposed to seal the cables water-tight. 

Mr. Robert Mallet, C.E. (Dublin), made a report to the British 
Association in 1858 on paints for bridge and cable iron work: "That 
he had tested ten of the best and most reliable ferric paints and var- 
nishes then known, and not one of them remained adherent and 
undecomposed for a single year under water. In moist air, and 
under conditions resembling English sea-coast fog, their state was 
not much better. The presence of moisture even to the extent of 
a partial saturation of the air, developed a fungus, the decomposition 
of which was almost as fatal to the life of a paint as immersion in 
sea-water." 

Government authorities state that there are over 300 suspen- 
sion bridges in Europe, of a great variety of spans and industrial 
importance; many of them having eyebars instead of wire cable 
suspensions. It is also stated that the life of the wire cables and 
anchorages have been found to be precarious, for oxidation was 
in progress in the interior of the cables, while the anchorages were 
weakened from the attack of some element "not at present defined " 
(evidently electrolysis). That there was no reliance to be placed on 
the preservation methods, or any certainty that they had effei't on 
the Ufe of the structure beyond twenty-five years. 

The failure of the Anglers wire cable suspension bridge showed 
that it was impossible to keep the hydrate-of-lime coating used there 
in immediate contact with the anchorage metal. Moisture and 
earth acids, carbonic acid from the atmosphere reached the metal, 
and the lime-coating was practically useless to prevent corrosion. 

M. Bemadeau in the "Annales des Fonts et Chausseur," 1881, 
refers to a bridge in which of the 150 wires forming a cable only 16 
were in good condition, the rest were brittle as glass. The bridge 
had been in use less than forty years. 

Two other suspension bridges of short span fell after twenty-six 
and twenty-eight years' duty. 

The suspension bridge over the Ostrawitza River at Mahrisch- 
Ostra, finished in 1851, failed in 1886 from a fracture of one of the 
anchor chains. The metal in the chain had thoroughly changed its 
character by corrosion, so that it could be crushed by the hand. 



CORROSION IN SUSPENSION BRIDGES, 391 

The anchor-bars in this bridge consisted of 12 links, one of which 
was completely corroded away and the others were reduced to about 
oneHsixth of the original size. The original sectional area of the anchors 
was 24.4 square inches, but had corroded to about 4 inches. An 
official examination and report of the strength and condition of the 
bridge was made in 1885, one year before its failure, which stated 
that ''The bridge has been examined in all its parts and is in good 
and safe condition.'' A squadron of Uhlans went down with the 
bridge. 

An examination of the wire cables of a suspension bridge, where 
coal-tar and lime had been used to coat the wires, also to fill the 
interstices between them where the cables entered the anchorages, 
showed that these cables were wrapped with ^-inch diameter wire 
and then a canvas jacket satiu*ated with coal-tar and lime placed over 
them. After less than twenty years' duty the tar had partially 
decomposed and disappeared and the cavities were filled with a 
dirty, grayish liquid. The wrapping wire, also the seizing wire on 
the strands, were in many cases rusted through, and the cable wires 
deeply pitted. The damage to all the wires extended about three 
feet upward and outward from the cable anchorages. Beyond this 
there was a little rust, but no pitting, and still further from the anchor- 
age the paint on the interior of the cable was gummy and undried. 

* French engineers of reputation now prohibit the use of white- 
lead or any quick-drying paints on anchorage cables. The failure of a 
number of suspension-bridge cables in France was directly traceable to 
the use of that kind of paint. Chalking and cracking of the coating, 
owing to the ceaseless changes of temperature to which they and the 
cables were exposed, admitted water and held it, and corrosion of the 
wires at or near their lowest position in the cables was the result. 

Euphorbium paints possessing elasticity, tenacity, and a quality 
that prevents them from drying bone-hard and becoming brittle, 
have proven the best paints used by French engineers for cable or 
other ferric work. Euphorbium being of a non-corrosive and anti- 
fouling nature, prevents the growth of atmospheric fungus, the 
decomposition of which produces an acid highly corrosive to iron. 

Other resinous or varnish paints dried hard and brittle and soon 
cracked. Mastic proved to be the best of all the so-called copal 
gums for a bridge-cable varnish paint. 

♦ Le Gdnie Qvil, 1881. 



392 CORROSION IN SUSPENSION BRIDGES, 

The failure of so many wire cable suspension bridges that have 
been in duty less than fifty years whose collapse can be attributed 
to corrosion or electrolysis and not to overloading, shows the imi>era- 
tive necessity of having the cable and particularly the anchorage 
metal accessible for inspection at any time. Corrosion of cables and 
metals in the air may be in part observed, but electrolysis occurring 
in the lower and hidden parts of the anchorage, and not reparable 
nor preventable, should be guarded against by a metallic connection 
at the anchor-plate end, that will lead off all electric currents and 
extinguish them in the earth and not in the metal of the structure. 
This connection can be renewed when corroded, and all shunts or 
attempted cut-offs of the current above the groimd line avoided. 
In the case of divided currents, cut-offs have been found to be un- 
reliable. 

Cement coatings or concrete cannot retard electrolysis if any 
moisture is present. Mr. Eiffel found the iron rag-bolts placed in 
fortification masonry two hundred years ago, had enlarged from 
two to two and a half times their original diameter by rusting in 
the mortar in a dry location. 

Rust once established, carries within itself the elements for a 
ceaseless life. Even in a glass bottle rust begets rust. Hydrated 
rust carries over 20 per cent of moisture, and so long as it can attack 
a fresh surface of iron and cast off the thin film of oxide as it forms, 
it will release enough oxygen to begin another cycle of action. 

Farraday's law of the corrosion of metal in weak acidulated 
solutions and electrical energy applied to the anode is 1.0448 grains 
of iron per square foot, per ampere hour. The life of any ferric body 
can thus be approximately ascertained before its construction. The 
methods and means for its preservation should receive the most care- 
ful consideration of the engineer and others responsible for its pro- 
tection. That its preservation is more difficult than its planning 
or construction does not remove this serious responsibility. 

The borough of Brooklyn (New York City) has about 800 miles 
of water-mains of all diameters from 4 to 48 inches. Many of these 
pipes in the early construction of the watei^works were cast in Glas- 
gow from a firm close-grained cast iron which showed a white sur- 
face on fractiu'e, indicating a large amount of combined carbon in 
the metal. Many miles of these pipes were also coated with Dr. 
Angus Smith's anti-corrosive compound, as before mentioned in 
this work. Previous to the year 1900 but few cases of electrolysis 



ELECTROLYSIS OF UNDERGROUND METAL. 393 

were reported in the water-mains of this city. The apparent freedom 
from electrolysis was attributed to the sandy soil in which the pipes 
were laid, also to the composition of the cast iron- 
Scotch gray or forge iron was thought to be exempt from corro- 
sion, and the Roebling bridge anchor-plates were made from this 
brand of cast iron and placed in the anchorages imder this idea. 
But many of the pipes were cast in American foundries from American 
cast iron, and but little difference in their corrosion and that of the 
Glasgow-made pipes had ever been noticed. After the effect of cor- 
rosion by electrolysis had been noticed, to determine whether the 
composition of the cast iron had any power to prevent it, pieces were 
cut from the foreign and American cast-iron pipes, also from soft 
cast iron containing but little combined carbon and more graphite 
than the Scotch irons, and used as anodes in various electrolytic cells. 
The electrolytes consisted of samples of earth from various parts of 
the city, moistened with distilled, hydrant, and sea-water. The cells 
were exposed to the action of currents of different voltage and 
amperage. 

In every case the anode was corroded, showing conclusively that 
there is no immunity from electrolysis of cast iron used for uxjUer-pipes, 
because of its chemical composition. 

It was determined by the observation of the water-works' engi- 
neers that the tubercular corrosion on the water-pipes when unpro- 
tected other than by the usual thin coal-tar or bitumen pipe dips 
was at the rate of about -nAnr of an inch yearly, there being a differ- 
ence in the rate of corrosion in the pipes of different metals, markedly 
in favor of the close-grained, white firm irons. 

Summarizing the report of many other water-works' engineers, 
it appears that corrosion from electrolysis, tubercules, or from other 
causes is more rapid in wrought-iron than in cast-iron pipes, irre- 
spective of the kind of soils they are buried in or to whatever influences 
they may be exposed. 

The small amount of electrolysis in the city of Brooklyn gas- 
and water-mains was finally attributed by Prof. Samuel Sheldon* 
to the presence of the hard, thin, vitreous scale formed on them at 
the moment of casting in green sand molds, and that this scale was 
s, non-conductor of electricity. This scale is similar to that noted on 

* Brooklyn Pol Inst Trans. American Institute of Electrical Engineers, 
May, 1900. 



394 ELECTROLYSIS OF UNDERGROUND METAL. 

car- wheels and in connection with mining-pipes; it has been fotind 
to retard corrosion due to sulphureted water, and has been referred 
to on page 328. 

A piece of the sand-coated pipe was covered with an insulating 
paint, but leaving exposed a small area of the silicate coating. This 
pipe was made the anode in an electrolytic solution. Less current 
flowed through the solution under a given E.M.F. than under sin:iilar 
conditions with an anode from the same piece of pipe, but exp)osing 
a clean iron surface of the same area to the same currents. In 
some of the experiments, no current at all passed the scale until 
the voltage was raised to a number of ohms. The water-pipes coated 
with Dr. Angus Smith's compound appeared to be less affected by 
electrolysis than the pipes not coated. The insulating quality of 
the compound added to the power of the silicate coating to resist the 
stray electrical currents of low potential. In certain districts of 
the city where high potential currents reached the pipes, electrolysis 
was present, but concealed by the firm and unbroken coating of 
Dr. Smith's and other heavy anti-corrosive coatings and was the 
more dangerous on this account. 

In every city in which electric street railway, lighting, and power 
service is developed, there will be a number of districts corresponding 
to the number of power stations and plants for individual electrical 
generation. Each one of these installations will draw currents from 
its ow^n district that can generally be very closely defined in the 
ordinary working of the station. But these boundaries become verj- 
irre^^ular, daily and hourly, from the varying nature of the currents 
required for the work to be done in them respectively. A difference 
of potential has been noted — as high as 40 volts between different 
points in the same district or between adjoining districts. Hence 
these boundaries are always shifting to a greater or less extent at all 
times. It matters but little from which district the current reaches 
underground metal or what its potential, electrolysis is assured in 
every case. 

Where electrical installations of a known amperage of 40,000 
to 50,000 are in daily use, there will inevitably be some leakage of 
the direct and return currents as well as a certain energy in the induc- 
tion currents, always present for the corrosion of metal, whether 
under ground or partly under ground and partly in the air. Perfect 
insulation from one or more or all of these currents is absolutely 
impossible. The best practice as it exists at present can only min- 



ELECTROLYSIS OF UNDERGROUND METAL 395 

imize their effect. While the writer has no desire to appear as an 
alarmist, the plain facts may as well be recognized now as hereafter, 
when the particularly dangerous character of stray electrical cur- 
rents of low voltage and large amperage is forcibly presented to the 
public in the sudden collapse of some important structure or gas- 
and water-supply systems. 

It is a false reliance that masonry, mortar, concrete, or cement 
are impervious to moisture and incapable of acting as an electrolyte 
such as would induce electrolysis. They are not insulating sub- 
stances, or at the best only in the smallest degree under the most 
favorable circumstances. They are positively porous and in nearly 
every case, whether tested in large or small mass, are permeable to 
all waters or moisture and gases, and in but a few exceptional cases 
ever become thoroughly dry. 

* A number of electric-light cable conduits in Paris were constructed 
of concrete, particular care being exercised in the selection of the 
hydraulic cement and sand used, as well as the ranuning of it into 
place. The conduits were far above the water-line of the city's soil 
and were considered to be water-tight. The copper wires soon 
became covered with verdigris and copper chloride and so reduced 
in area that grounding and heating were of frequent occurrence from 
the normal currents of the service. A number of minor explosions also 
occurred, due to the gases formed by the decomposition of salt 
water that filtered through the cement when salt was strewn on the 
roadway over the cable conduit to melt the snow. The gaseous mix- 
ture contained oxygen, hydrogen, and chlorine, the latter gas being 
due to the chloride of sodium in the salt water. The leakage of the 
current furnished the electrical energy to decompose the salt water 
and fire the mixture, also to form the carbonate of soda and caustic 
soda that was deposited upon the copper wires. 

Earthenware conduits were also used, but were not water-tight, 
and the same decomposition of the salt water and corrosion of the cop- 
per wires occurred as in the concrete construction, and their use was 
soon abandoned. The electric cable wires are now taped or covered 
with a bituminous compound to prevent electrolytic action. 

President Learned, in his inaugural address to the New England 
Association of Gas Engineers, March meeting, 1902,t stated that 



* LElectricien, 1892. 

t ''ElectrolyBis of Gas Pipes." American Gas Light Journal, March 3, 1902. 



396 ELECTROLYSIS OF NEW ENGLAND GAS PIPES. 

the conclusions derived from a large number of tests and observations 
of the effect of electrolysis on the gas-pipe systems in a number of 
cities in New England were: " That gas-pipes laid with lead joints 
have 15 per cent greater resistance to electric currents than water- 
pipes of the same diameter with similar joints. Screwed joints in 
wrought-iron pipes have about the same resistance as the lead joints 
in cast-iron pipes of the same diameter. The resistance of a Port- 
land-cement joint, as ordinarily made, was from 15,000 to 20,000 
times the resistance of a lead-caulked joint, comparing pipes of equal 
diameter. That the resistance of the cement joint depended in a great 
measure upon the amount of moisture that the cement takes up after 
setting. 

" The conclusions drawn from the experiments made on gas-pipes 
of all diameters laid in short or long sections were: That all possible 
resistance should be inserted in the pipe mains by making the joints 
of some insulating or semi-insulating material, with an asbestos or 
tar-paper ring between the abutting ends of the pipe in order to 
break up the pipe-line into as many metallic units as possible, and 
isolate them from all other pipe or trolley systems, so far as practi- 
cable and mechanical conditions would allow. The pipes should be 
coated with a water-proof compound. The ordinary foimdrj^-dip 
coating or painting with oil-paints had no appreciable effect to delay 
or diminish electrolysis of the pipes. Neat hydraulic cement cover- 
ings were worthless on account of the porous nature of the cement. 
It absorbed moisture, was inelastic, and easily scaled off the pipes by 
frost or mechanical injury. 

*' A covering made from 3 parts of dry clean sand and 2 parts of 
coal-tar boiled to a pitch at 660° F. made a mixture that was slightly 
elastic at ordinary temperatures, was thoroughly water-proof and 
when applied to the pipes to a thickness of 1^ inches, had an insulat- 
ing resistance of over 1 million of ohms to the cubic inch. A short 
piece of 2-inch pipe covered with this mixture and inrnnersed in a 
strong solution of water and soda ash for 8 hours, showed no signs 
of the absorption of any of the solution, nor had any decrease of 
electrical resistance been developed." 

Mr. Robert Irvine, F.C.S., reports to the Chemical Society "that 
the cause of many disastrous explosions from the leakage of gas had 
been found to be due to electrolysis between the brass unions and 
other composition fittings and the iron service pipes and casings of 
the meter. Voltages between the brass and iron materials ranged 



ELECTROLYSIS OF EAST RIVER SUSPENSION BRIDGES. 397 

from .3 to .5 of a volt, the iron in all cases being the electro-positive 
metal." 

The corrosion by atmospheric exposure in the Brooklyn Sus- 
pension Bridge superstructure is deeply seated in every square foot 
of the structure and is beyond correction except by rebuilding it. It 
takes a corps of painters constantly at work three years to paint 
the structure, and the coating principally serves to mask the corrosion. 
The voltage of the electrical currents passing through the suspension 
cables has been referred to, but their corrosive effects upon the an- 
chorage metal from the inaccessibility of the lower ends must always 
be a conjecture. These parts are beyond repair, and the electrical 
currents, whether from the trolley railway on the bridge, or from 
the scores of large installations surrounding the structure, are of 
large amperage, constant and uncontrollable in action. That they 
will not prove active agents of electrolysis is not in accordance with 
past experience. 

In the other East River suspension bridges, where steel instead 
of masonry piers are used to carry the suspension cables and the 
superstructure, it is expected that the large metallic contact of the 
wire cables at the top of the steel towers will form a short circuit 
and ground for any electrical energy that may reach them from the 
railway, instead of using the anchorage metal for a terminal. This 
theory can only be determined after the bridge railway has been 
put in operation. The many points of junction of the bridge trestles 
with the masonry piers on both sides of the river may divide the 
trolley currents into a number of short circuits so as, in a measure, to 
protect the structure. But this cannot prevent the currents that 
come to the anchorage metal from the installations surrounding them, 
from using the cables for their transmission, as they provide the 
best and shortest metallic path for their ceaseless circuit. 

A further fact relative to the electrolysis of the anchorage metal 
lies in its condition, even before any material strain other than the 
weight of the foot-path cables came to it. In the Williamsburg 
Bridge, the anchorage pits were carried down into the solid rock and 
near if not below the water level of the river. These were not sealed 
water-tight by any effectual method in laying the superincumbent 
masonry. The anchorage metal was put in place under a continual 
seepage of brackish water or that rendered alkaline from the cement, 
more or less pumping being required during the work. The anchor- 
age metal, also the forged eyebars for the cable connections were 



398 ELECTROLYSIS IN WILLIAMSBURG SUSPENSION BRIDGE. 

indifferently cleaned, instead of being sand-blasted, and were painted 
with Smith & Co/s "Durable Coating," the advisability of apply- 
ing a baked japan coating being disproved on account of its cost. 

When the chain of eyebars were ready for the cable construction, 
an anchorage pit was pumped dry and the eyebars inspected. Though 
the bars were covered with two coats of "Durable Coating," and in 
place only about two years, the paint was nearly destroyed, and corro- 
sion over the whole surface of the bars wherever the water had 
reached them was virulent. The limited room in the pit and the close 
association of the eyebars together and to their bed, rendered the 
cleaning and repainting of them difficult and in many feet of their 
length impossible. A new coat of varnish paint, however, was 
applied in the damp atmosphere of the pit. 

For the future protection of the eyebars it is proposed that when 
the bridge is completed and the chain of bars are bearing their load 
and have adjusted themselves to their permanent position, to paint 
them again and fill in between and around them for a foot or more 
with melted bitumen and to fill the pit with concrete. This plan does 
not reach the anchorage plates and metal in the lower end and inac- 
cessible part of the pit. The corrosion serpent is only scotched (not 
killed) and will remain inactive for but a short time, or only so long as 
the concrete remains thoroughly dry, something impossible to maintain. 

The bitumen coating insulates the eyebars so far as it can be 
applied and passes on whatever electrical currents may reach them 
from any source to the lower end of the anchorage, where the metal 
is not protected, corrosion in progress, and inspection almost if not 
quite impossible. To think that electrolysis will not take place in 
the metal at both ends of the bridge is to ignore facts already 
established. Electrolysis, or even corrosion from the contact of 
metal with moisture, in this case, is not the sin of the paint manufac- 
turer; where to place the blame is not hard to find. 

The introduction into marine service of appliances for the genera- 
tion and use of electric power and light has developed a new field for 
electrolysis, that seriously endangers the efficiency and life of all vessels 
so equipped. An examination of the United States cruiser Brooklyn, 
for the purpose of determining the effects of a recent grounding of the 
vessel, has revealed the fact that electrolysis has attacked the inner 
skin or false bottom of the ship, and it is in such an advanced state 
of corrosion as to be practically destroyed. That she survived the 
grounding accident is a cause of wonder to the naval officers. 



ELECTROLYSIS IN MARINE CONSTRUCTION. 399 

Electrolysis on board a steel ship is not unlike the same develop- 
ment by direct or stray electric currents in land or underground struc- 
tures. Wherever a current of any potential leaves the metal, elec- 
trolysis is the result. In the case of naval vessels there is an enor- 
mous amperage present at all times and that cannot be returned to 
the dynamos, even with an increased capacity of the return-current 
wires over those employed for distribution of the current. 

Divided and induced currents, also the electric enei^gy developed 
by corrosion itself, will seek their own course either in returning to 
the d3niamo or extinguishment in the ground connection. The 
latter, in the case of marine work, being water saline or foul in char- 
acter, is a more efficient electrolyte than any earthy substance. 
Hence electrolysis in marine constructions willl naturally be more 
rapid and virulent than on a similar ferric area and current exposure 
on land or in underground structures. 

As before stated, no paint or plastic coatings of the metal will 
prevent electrolysis. At best they may temporarily mask its progress, 
but it only requires a short time or a slight change in the conditions 
to reveal it. 

Corrosion or electrolysis of marine metal can only be controlled 
by the use of some alloy of steel that will minimize the action, and 
by such an increase in the thickness of the parts of the ship exposed 
to corrosive influences as will for a time provide for any reduction 
of strength in the corroded parts or alloyed metal; also by a plan of 
construction that recognizes the possibility of the evil and provides 
that corroded members can be removed without practically rebuild- 
ing the ship. (See page 339 for Dr. Wurtz's protective method.) 

Insulation of motors and their connections and the positive pro- 
hibition of the connection of any electric current, even of the smallest 
amount, to any part of the structure will reduce but not prevent 
the dangers of electrolysis, whether in marine or land constructions. 

In either case, constant and thorough inspection of all ferric 
surfaces by an inspector who knows what to look for, and is compe- 
tent to recognize it when it is found, is an essential. Even if the 
inspector cannot avert corrosion when found, at least the danger 
can be noted, watched, and a warning given when it is time to desert 
the ship. All of these essentials appear to have been absent in the 
case of the cruiser Brooklyn, and probably they are also neglected 
in all of the steel vessels in commission. 



CHAPTER XXXV. 

ANTI-CORROSIVE MARINE PAINTS AND ALLOYS. 

The so-called marine paints are those applied to ships' bottoms 
to prevent corrosion, also those to prevent the growth of maiine 
plants, barnacles, etc., and known as anti-corrosive and anti-fouling 
paints. The anti-corrosive marine paint is not applicable to ferric 
bodies in any other place than under water, as in the air they crack, 
flake, or crumble off rapidly. They draw the necessary oxygen to 
dry them from the water and carry lai^e quantities of metallic salts 
and volatile driers to enable them to harden in an hour or so, an 
indispensable quality in a paint for a ship's bottom. One of the 
earliest patents in the arts was issued in 1670 for a tar and asphalt 
varnish for ships' bottoms, and since that time patents for marine 
paints have been issued, experiments and extended applications of 
them have been made by the thousand, and yet neither the corro- 
sion nor the fouling of ships' metal has been rendered materially 
less than it was a hundred years ago. 

Fossil resin varnishes prove to be the best vehicles for marine 
paints, whatever the composition of the pigments assembled with 
them. The nature of the pigments in these paints has a governing 
influence in exciting the galvanic action between themselves and 
the ship's metal, which action is speedily fatal to the coating as well 
as rapidly increasing the corrosion. 

Corrosion in marine constructions is not only increased by the 
action of sea-water on the pigments and covered metal, but by the 
great porosity of the paints on account of the quantity of volatiles 
used. This porosity of the coating is also a prime factor in the decay 
of paint on land structures.* 

The following tabulated results of a test of a large number of anti- 
corrosive coatings exposed to sea-water shows how inefficient nearly 
all of them are to meet the requirements of marine exposures. The 
baked coatings, Nos. 158, 159, 174, 175, 35, 113, 104, 105, 122, 
124, not being applicable for a ship's bottom, however effective they 

* Pages 405, 406, 407. 

400 



MARINE ANTI-CORROSIVE COATINGS, 401 

may be for many other marine purposes, are practically eliminated 
for comparison with the other marine paints; but are useful to com- 
pare with the other coatings if used on land exposures where the 
corrosive influences are not so severe. Porosity in the baked coat- 
ings is eliminated and corrosion lessened whether the coatings are 
used on land or for sea-water exposure. 

The Kauri and Zanzibar varnishes, composed of 20-30 or 40 
gallons of linseed-oil to 100 pounds of fossil resin, were not as effec- 
tive for preventing corrosion as where a pigment was added 
to the same quality of varnish, the order of merit for the pigments 
so used being carbon, zinc oxide, graphite, red lead, and iron oxide. 

The test, while of value as a record of the comparative merits of 
the several coatings to resist corrosion, when exposed on a small plate 
to sea-water under absolutely uniform conditions for all of the coat- 
ings, still lacks the factor of their behavior when applied in mass of 
material to a ship or to land structures exposed to sea-air or sea- 
water. 

The corrosion in the aluminum plates indicates that the metal 
to be non-corrosive must be alloyed with a metal lower in the electro- 
chemical scale than copper, in order to render aluminum of any 
practical value for constructions of any magnitude where strength 
or permanency are required. 

See Mr. R. P. Hobson's (naval constructor, U. S. Navy) report 
to the U. S. Navy Department "On the Uses of Aluminum for Naval 
Work," published by the Navy Department, Washington, D. C, for 
other data on the corrosion of aluminum. 

In some experiments of the "Alloys Research Committee" by 
Prof. Roberts- Austen, F.R.S., "samples of alloys containing 40 to 
60 per cent of aluminum were kept a number of months before being 
analyzed. During this period they had spontaneously disintegrated 
to a powder. The powder was not oxidized, but consisted of clean 
metallic grains, probably resulting from chemical changes which had 
taken place in the solid alloys. Whether the iron and aluminum were 
in a state of solution, or were chemically combined when molten, 
they are evidently chemically combined in the metallic powder as 
attempts to melt it are unsuccessful, which indicates the formation of 
an infusible compound. The two metals may have been too hot to 
unite thoroughly when in a molten state, but a long-continued 
proximity at a lower temperature aflFected their chemical union." 

M. Le Chatalier, in a paper read before the Acad6mie de Sciences 



402 CORROSION OF FERRIC ALLOYS. 

Paris, stated that equal parts of aluminum and copper were fused 
together in a crucible. The ingot was placed in a solution of common 
salt and lead chloride for twenty-four hours with a view of dissolving 
out the imcombined aluminum. No change in the ingot was apparent at 
the end of this time. The ingot was removed from the bath, washed, 
and dried. Twelve hours afterwards the ingot was found to be 
reduced to powder from the spontaneous oxidation cf the alloy. A 
similar ingot not immersed in the saline bath was unchanged at the 
end of a month. 

Three small aluminum boats, used by Mr. Wellman in his 1894 
polar expedition, soon after they were brought back could be crum- 
bled in the hand. 

Corrosion of Ferric Alloys. 

The influence of copper and nickel alloyed with wrought iron and 
soft steel was made the basis of a paper by Mr. F. H. Williams, C.E., 
read before the Engineering Association of Western Pennsylvania. 
The paper was based upon experiments made in the line of 
some recent investigations by Mr. H. M. Howe, as given in his paper, 
"Relative Corrosion of Wrought Iron and Soft Steel and Nickel 
Steel," read before the International Congress, Paris, on Testing 
Materials.* 

"In prosecuting the tests, Mr. Williams selected four samples of 
steel, viz.: A, an ordinary soft Bessemer steel; B, C, D, soft Besse- 
mer steels to which copper had been added in the converter so that 
they contained respectively 0.078, 0.145, 0.263 per cent. In addi- 
tion, another set of test materials, consisting of one soft Bessemer- 
steel sample and four of wrought iron, were similarly alloyed, sample 
number 4 of wrought iron having 0.393 per cent of copper. All 
the samples were brought to the same dimensions, then weighed 
and suspended in a frame, so that they could all be dipped simul- 
taneously in water and left to dry in the open air, this treatment 
being repeated frequently each day for a month. The daily increase 
in weight due to oxidation was small, but of such a persistent char- 
acter as to indicate the retarding influences of the copper upon the 
corrosion. Finally, where there appeared a tendency of the oxide 
to scale off, the treatment was suspended, the samples dried, cleaned 
from all oxide, and weighed. The loss in the original weight of the 
samples is tabulated, viz.: 



* Engineering Record, December 1, 1900, p. 519. 



CORROSION OF FERRIC ALLOYS. 403 

Loss FROM One Month's Exposure to Atmospheric Corrosion. 

A. Soft Bessemer steeL 1 .85 percent. 

B. " " " with 0.078% copper. 0.89 

G " " " " 0.145% " 0.76 

D. " " " " 0.263% " 0.74 

Soft Bessemer steel, second sample 1 .65 

Wiought-iron sample nimiber 1 with 0.078% copper 0.76 

" " " 2 " 0.145% " 0.80 

" " 3 " 0.263% " 0.87 

" " 4 "0.393% " 0.53 

" Mr. Howe's experiments indicated that in large plates of metal 
containing nickel in approximately the same proportions as the above 
examples, and exposed for a considerable time, the corrosion was simi- 
larly retarded. The introduction of a small amount of copper or 
nickel into soft steel can be easily effected, and their presence within 
the amount necessary to obtain the above results has been demon- 
strated not to be prejudicial to its physical properties. The data 
here presented, while not showing that corrosion of ferric bodies can 
be wholly prevented by alloying them with copper or nickel, they 
do indicate that a soft steel can be made capable of resisting corrosion 
quite as well as wrought iron and thus settle the debate about theii 
relative corrosibility now so much in question." 

In the case of aluminum, the natural field for its use appears to 
be in marine construction, where lightness is' an essential requirement, 
particularly in naval construction, where economy in weight has a 
vital relation to military efficiency. But its corrosibility, if exposed 
to sea-air or sea-water, has demonstrated its imfitness for any struc- 
ture exposed to these influences. 

Alloys of copper with aluminum increase the corrosion, which is 
greater as the amount of copper is increased. With 2 per cent of 
copper the increase of corrosion is markedly greater than with the 
pure aluminum, and where 5 per cent of copper is used, as in the 
case of the Yarrow torpedo-boats, on account of the great increase 
in the strength of the metal, the corrosive effects were disastrous, and 
caused an abandonment of the metal for French naval work. The 
interest in this feature is special, as the aluminum needs an alloy to 
increase its strength, and copper appears to be the one metal best 
suited for this purpose, though other metals are available. The 
increased corrosion due to the presence of copper is attributed to 
the fact that the two metals are so widely separated in the electro- 
chemical scale that an alloy made from them contains the necessary 



404 ACTION OF SEA-WATER ON METALS. 

elements for a strong gaivanic action that soon destroys the integrity 
of the metal. This action is also developed where copper in Bxiy 
fonn is brought into contact with iron or steel for any exposure. 

The action of sea-water on the corrosion of various metals has been 
investigated by the German engineer, Digel, who reports "that alloys 
of copper 20 per cent and nickel 42 per cent are not very rapidly 
corroded. Adjacent masses of iron, copper, or copper alloys render 
the copper, and nickel alloys somewhat immune against sea-water 
corrosion, the above-mentioned metals being rapidly corroded. 

" Copper and zinc alloys are corroded almost uniformly over the 
surfaces exposed to the sea-water. When the zinc exceeds 24 per 
cent of the alloy, it is leached out, leaving a brittle porous mass of 
copper. Adding 15 per cent of nickel to the alloy prevents this 
leaching action. 

"Very pure electrolytic copper, in contact* with ordinary com- 
mercial copper (99 per cent pure), is very rapidly corroded by sea- 
water. When the two coppers are not in contact, they corrode about 
alike. Both of the above brands of copper when annealed are more 
rapidly corroded than when rolled. Copper coated with zinc is 
temporarily protected from sea-water, but when the zinc has been 
dissolved, the corrosion of the copper is increased rapidly. 

" Electrolytic copper that had its surface oxidized in places devel- 
oped rapid galvanic action in the clean spots. Copper pipes brazed 
into vessels or into each other are subject to corrosion in the brazed 
joints from galvanic action. 

" One-half of 1 per cent (0.5) of arsenic in metal greatly retards 
corrosion. 

" Wrought iron and steel of various methods of manufacture are 
greatly influenced in corrosibility by the amoimt of phosphorus 
contained in them. With 1 per cent of phosphorus present the corro- 
sion was a little over one-half as great as in the case where the iron 
was free from phosphorus. When two pieces of phosphorus iron 
or steel were in contact, the sea-water corroded the low phosphorus 
metal from two to five times as fast as the high-grade metal. 

" Iron alloyed with nickel showed the same behavior. 

'* The normal corrosion of single plates of metal was less as the 
percentage of nickel increased. When two plates difTering in the 
contained nickel were brought into contact, the plate higher in nickd. 
was almost completely protected from corrosion." 



MARINE PAINT TEST 



405 






CD 'O 





CO 

a 
o 

a 



I 

a 



O 

e 



0) 



'C 

'Q <• *• <• ^ JJ 



•c 



,£: 



o 



N 



GR 

•c 

OS 



09 

08 



QQ 



? 



OS 

a 
a 
tsj 

o 

CO 



g5 

03 



03 â–º*< 



E 

08 
> 

m 

O 



o 

CO 



?^ 



£ 

OS 

ft* 

a. 

QQ 



C 

§ 
3 

« 

& 



08 
> 

c 

2 3 



tS 



CO ^ 



K 



ai 

Xi 

N 

c 

03 
S] 

O 
CO 



0> 



SSaSSSEjSSSSSSSgSfeSSgKS 



406 



MARINE PAINT TEST. 



CD 

Qi 

a) 



9 

i 

s 

I 



a 
o 



(t 
I 



o 



CO 

c 

Z 

M 

o 



«D 



gd § 

5 5 ■•*» 

*5 C «^ 
es eS ;3 

« a "-* 









g§ 

c o 

is O - 
oj o * 

ffiO 











o 

•ss 

CD 



^^ 

^^ • ■% 
QQ "O 

>*& 

no to 
Si no 



O 
0) 



e 



E 


-§ 


il 


00 


CO 


C 




o 


-Q 


u 


0) 


â–  an 

0) 


C; 


s 




o 


c 


t« 


o 


>» 




1 


c 


Xi 


8 


a 


TS 


â– *s 


o 


flS 


o 


o 


O O 



o 

i 

CO 

H 






a 






0) 



^ 



• 






• 


• 0) 


3 








ball 
oxi 


0) 


i 


0) 


g 


d lead. 
ince*s mel 
rple iron 




pC <• « 

{3 


- to - '• 




0) 


o; 


«? C 3 


5 


O 


w 


tf 


tf Ph £ 



> 8 



V QQ 

I 

o 
-♦a 



O A 



ec Tf ic 

b- b- h* 




CO 

c3 



o o o 

CvJ CO rf 



â– g 






<S 



03 
> 



•o S - 5 -s 



c 



_ J2 

•3 g 
2 I 



-♦a 

^ s 
- s 

CO S 






MARINE PAINT TEST. 



407 



M 



a -5 



OQ 

H 



si 

k 



2 . 



•111 

'3 9 8 



•SI'S 




S •• •• □ pH iT: ^ 'S ^^ 



? 



08 



05 



^ M5 r>. « 1-H w 
c3 & & R S S 



t^ 00 o» 



8 8 8 CO 



-*^S28g3S§ SSSJgS 



CHAPTER XXXVI. 



MISCELLANEOUS TABLES AND DATA. 



Pigments and Inert Substances. 



Substance. 



Asphaltum 

Alumina oxide (clay) 

Anthracite coal 

Bituminous coal 

Bitumen 

Barytes (heavy spar) 

Brick-dust 

Cement, hydraulic, common . . 

Portland. . 
" " Rosendale. 

Charcoal in bulk 

oak 

Coke, natural, Virginia 

hard foundry 

gas retort 

Cobalt (blue) 

Coal-tar (gas retort) 

Chalk, red and black 

" whiting, Spanish white 

Clay (see Terra- Alba) 

Carbon (diamond) 

Feldspar 

Flint 



Ssonbol. 



AlO 



BaSO, 



it 



(f 



Gneiss and granite 
Glass 



Graphite, amorphous 

" flake or foliated 

Gypsum, native sulphate of lime (hydrated) 

" (calcined) plaster of Paris 

Lampblack 

Lime (quick) 

Limestone, common gray 

" Carrara marble 

Lead, metallic 

white, carbonate 

chromate , 

hydrate 

sublimed 

sulphite (dark color) 

sulphate (white color) 

Lithai^e 



C 

n 
tt 
u 
It 



CaO 



n 
it 
tt 
u 
(( 



CaOSO,+HjO 

CaOSOa 

C 

CaO 

CaCO, 



tt 



Pb 

PbCO, 

PbCrO, 

Pb(OH), 

PbSO, 

PbS 

PbSO, 

PbO 



Specific Gravity. 



04 



1.8—1.39—1 

2.75—2.6 

1.70—1.35 

1.318—1.277 

1.16—0.83 

4.7—4.3 

1.50—1.30 

1.6—1.5 

1.51—1.25 

1.00—0.96 

0.441 

0.336—0.331 

0.746 

0.800 

0.70 

8.6-S.6 

1.1—1.00 

2.8—2.2 

2.1—1.8 

2.71—2.56—1.93 

3.529—3.65 

2.^—2.5 

1.93 

2.8—2.76—2.62 

2.782—2.5 

2.768—2.208 

1.40—1.21 

2.5—2.38 

2.4—2.08 

0.44—0.80 

0.88—0.8 

2.7—2.6 

2.717 

11.44—11.07 

6.480—6.465 

5.2—4.61 

7.00—6.6 

6.258 

6.43 

7.13 

9!00— 8.60 



408 



PIGMENTS AND INERT SUBSTANCES. 



409 



Pigments and Inert Substances — Continued. 



Substance. 



Lithopone (zinc sulphate). 

Red lead (minium) 

Marl 



Mica 

Masonry, brick or tile 

Magnesia, carfc>onate 

Manganese dioxide or p3rrolusite, 



Ochre. 



Iron, metallic 

" oxide (red rust) , 70% iron, 30% oxygen. 
Iron ore, red and brown; hematite, specular 

and columnar; spathic, etc 

Black magnetic oxide, 72.413% iron, 

27.587% oxygen 

Phosphate of lune (basic steel furnace slag) . . 



Pitch. 

Quartz 

Resin 

Orange or Paris red , 
Sand from quartz. . 
Silica (crystalline). . 
" (amorphous). . 

Silex (floated) 

Slate 



Symbol. 



PbSO, 

pb,o; 



» 



Fe 
Fe,0, 

Fe,0. j 

Fe,0« j 
Ca,(POJ, 



Specific Gravity. 



(C^OJ. 



Slag, furnace. 

Soapstone or steatite 

Sulphate of iron (copperas) 

Sulphur, native ore, melts at 116° to 120*» F. . 

Spanish brown (oxide of iron and clay) 

Umber (argillaceous brown hematite iron ore) . 
Terra alba, China, pipe and potters' clay, 

complex oxides of all metals 

Yermifion (mercuric sulphate) 

Vermilionette (mixed color) 

2inc metallic 

" carbonate 

" oxide or zinc white 

sulphide (dark color) 

sulphate (white color) 



tt 



tt 



SiSO, 
SiSO, 



Fe^O^ 



2Fei083i02+HsO 



! 



HgS 



{ 



Zn 

ZnCO, 

ZnO 

ZnS 

ZnSO, 



4.2 

9.07—8.94—8.06 

2.4—1.73 

3.1-2.76 

2.0—1.79 

3.2—2.72—2.4 

4.97—4.816 

4.6-3.2 

7.84—7.77 

6.77 

6.334.62 

4.20—3.80 

6.062 

3.6--3.8 

1.162 

2.7—2.64 

1.089—1.07 

8.1—7.6 

1.76—1.44 

2.8—2.6—1.9 

2.34—1.9 

2.8—2.3—1.8 

2.9—2.7 

3.6—4.3 

2.8—2.65 

6.176.04 

2.03—3—2.05 

2.2—2.1 

2.4—2.2 

2.71—2.66 

2.4—1.92 

8.91-8.1 

7.6—7.0 

7.0 

4.4—4.0 

5.6—6.4 

4.2—3.9 

6.3—6.0 



The trade or commercial color pigments number 245, viz. : 



Black 22 

Blue 28 

Brown 24 

Gray 3 



Green 46 

Red 60 

White 36 

Yellow 36 



77 



168 



410 



PIGMENTS. METALLIC BASES 



Characteristics of Metallic Bases of Pigments. 



Metal. 



Antimony 

Aluminum 

Arsenic 

Barium 

Bismuth 

Calcium 

Cobalt 

Copper 

Iron 

Lead 

Magnesium 

Manganese 

Mercury 

Nickel 

Potassium 

Phosphorus 

Silicon 

Sodium 

Sulphur 

Gold 

Silver 

Tin 

Zinc 

Red and brown hematite and specular 
iron ore; iron=70%; oxygen=30%. . 

Magnetic or black oxide of iron; iron= 
72.413%; oxygen =27.687% 



Symbol. 



Sb 
Al 
As 
Ba 

Bi 

Ca 
Co 
Cu 
Fe 
Pb 
Mg 
Mn 
Hg 
Ni 
K 
P 
Si 
Na 
S 
Au 

Ag 
Sa 
Zn 

Fe,0, 



Fe30, 



Combining 
Weight. 



i 



120.4 
27.1 
75.1 

137.4 

200.0 ) 

208.1 J 
40.1 
59. 
63.6 
56. 

206.9 

12. 

55. 
200. 

58.3 

39. 

31. 

28.4 

23. 

32,1 

197.2 

107.7 

119-117.8 

65.4 

56 and 16 



56 and 16 



Speeifio Grsrity. 



6.86 — 6.36 
2.71 — ^2.56 
6.96 — 5.62 
6.85 

9.90—9.74 

1.58 

8.6 — 8.6 

8.92 — 8.69 

7.77 
11.38 

1.743 

4.97 — 4.82 
13.62—13.58 

8.93—8.28 

0.865 

1.823—1.777 

2.8-2.5 

2.05 — 2.033 

19.36 — 19.245 

10.61 — 10.474 

7. 409-7. 300 

7.13 — 7.0 

5.4-5.3 
4.65-4.2 

5.6—5.3 



ELEMENTS THAT CAUSE THE DECAY OF PAINT. 411 



Gases and Elements that Cause the Decay of Paint, 

The principal gases and elements that affect all paints at a tem- 
perature of 60° F. and under one atmospheric pressure are: 



Substances, Gas or Vapor. 



Alcoholic vapor 

Ammonia NH, 

Atmospheric air 

'^ '' saturated as fog at 80^ F.... 

Benzine vapor, ^J^t 

Bisulphide of carbon, CS, 

Carbonic oxide, CO 

Carbonic-acid gas, CO,. 

Ethelene or olefiant gas, C^H^ 

Hydrocarbon (illuminating gas), CH,. 

Natural gas 

Nitrogen, N 

Oxygen, O 

Hydrogen, H, 16 times lighter than air ) 

14i " " " oxygen f 

Marsh gas, C^. 

Phosphoric acid, 3H0P0. 
Phosphorated hydrogen, FH, 

Steam, saturated, 212*» F 

'' gaseous 

Turpentme vapor 

Smoke of bituminous coaL 

" " coke 

" " wood 

Sulphurous-acid gas, SO.. 

Sulphuretted hyofrogen, H^ 

Sulphuric ether 

Wood-alcohol vapor, CH^O, 






Specific 


Number of Cubic Feet 


Gravity. 


per Pound. 


1.589 


8.27 


0.607 


20.95 


1.00 


13.11—13.14 


1.0236 


12.80—12.831 


2.7 


4.78 


2.6447 


4.86 


0.9727 


13.67—13.60 


1.629 


8.694 


0.9784 


12.680 


0.302 


33.112 


0.64 


24.336 


0.9714 


12.762 


1.066 


11.887 


0.06926 


189.662—189.70 


0.669 


23.479 


0.49969 


26.36 


0.622 


21.077 


2.76 


4.76 


0.102 




0.106 




0.09 




2.213 


5.90 


0.177 


74.422 


2.686 


5.08 


0.812 


boils at 140' to 160^ F. 



Oxygen in Pigments. 

Paint-trade literature bears so persistently upon the point that 
red lead, from the great amount of oxygen it contains is not only 
self-destructive, but will destroy all other paints of which it forms a 
part; also, that by reason of this inherent element, it acts the part 
of a carrier for an additional amount of oxygen that it may collect 
from other sources, the joint effect resulting in an early destruction 
of the coating; also in promoting corrosion of the covered surfaces. 
The following comparison of the amount of oxygen in a number of 
pigments and so-called inert substances in common use in paints is 
ef interest upon this point: 



412 



OXYGEN IN PIGMENTS. 



Substance. 



Litharge 

Red lead 

White lead (carbonate) 

Sublimated lead 

Sulphate of lead (native ore) . . 
" " " (pigment). . . 

Sulphite of lead 

Zinc oxide 

" sulphide 

" sulphate 

" caroonate 

Whiting, chalk 

Iron oxide 

" magnetic oxide 

Barytes 

Gypsum 

Manganese dioxide (pyrolusite). 
Vermilion 

Umber (hydrated), â– ! 



Symbol. 



PbO 

Pbfio, 
PbSO, 
PbSO, 
PbSO, 

PbS 

ZnO 

ZnS 
ZnSO^ 
ZnCOg 

CaO 

Fe^O^ 

BaSO, 

CaOSO, 

MnO, 

HgS 

2FeA.SiOj 

+ H,0 



MetaUic. 


Oxygen, 
Per <>nt. 


Other Elements, 


Per Gent. 


Per Cent. 


92.822 


7.178 




90.63 


9.37 




77 . 516 


17.987 


Carbon, 4.497 


71.831 


18.561 


Sulphur, 9.608 


72.09 


16.746 


11.164 


62.283 


21 . 145 


10.572 


86.67 


00.00 


13.43 


80.344 


19.656 




67.077 


00.00 


" 32.923 


40.495 


39.670 


" 19.835 


52.153 


38.278 


Carbon, 9.569 


71.479 


28.52 




70.00 


30.00 




72.413 


27.587 




65.172 
41.20 


22.097 
35.28 


Sulphur, 12.731 
^' 23.52 


43 . 161 


56.839 




86.207 


00.00 


" 13 . 793 


55.735 


34.67 


Silicon, 9.594 



Combinations of Oxygen with Carbon and Sulphur. 



Substance. 



Carbonic oxide 

Carbonic acid 

Carbon trioxide 

Sulphuric oxide 

Sulphurous acid (the acid of 

burning sulphur) 

Sulphuric acia 

" " anhydrous 



Symbol. 



CO 

C02 

C03 

so 

so, 
so, 
so. 



Oxjrsen, 
Per Cent. 



57.14 
72.73 
80.00 

33.264 

50.00 
60.00 
66.6 



Carbon, 
Per Cent. 



42.86 
27.27 
20.00 
Sulphur. 
67 . 736 

50.00 
40.00 
33.4 



Number of 

Cubic Feet 

of Gas in 

One Pound. 



13.57 
8.59 
6.47 



3.848 
3.569 
2.944 



Changes in Pigments Due to Atmospheric Influences. 



Red lead (Pb,04)=lead, 90.63 per cent; oxygen, 9.37 per cent. 
Specific gravity, 9.07. 682 grammes, volume— 75.2 c.c. 

Upon exposure to hydric-sulphide gas in the atmosphere it 

Changes to 
Red-lead sulphide (PbS) = lead, 86.61 per cent; sulphur, 13.39 per cent. 
Specific gravity, 7.13. 714 grammes, volume =100.1 c.c. 
Increase in volume, 24.9 per cent. 



CHANGES IN PIGMENTS. 413 

Ziiic oxide (ZdO)= zinc, 80.344 per cent; oxygen, 19.656 per cent. 
Specific gravity, 5.42. 81 grammes, volume = 14.9. 

Upon external atmospheric exposure, absorbs carbonic acid and 

Changes to 

Zinc carbonate (ZnCO,)=zinc, 52.153 per eent; ] ^"^^f^"' ^f '^I^ P^' "^'^*' 

â– ^ ^ M carbon, 9.569 per cent. 

Specific gravity, 4.44. 125 grammes, volume— 28. 

Increase in volume nearly double. 

White lead (PbCOg)=lead, 77.516 per cent; j oxygen, 17.987 per cent; 
Hydrated carbonate; ( carbon, 4.4972 per cent. 

Specific gravity, 6.480. 

Absorbs carbonic acid from the atmosphere and 

Changes to 

Subcarbonate of lead (PbCO,)=lead, 73.138 per cent; ] ^^y«®''' ^'^^ J^"" ''®"^' 

* '^ ' ( carbon, 6.242 percent. 

Specific gravity, 6.40. 

Absorbs sulphurous-acid gas and 

Changes to 

Sulphide of lead (PbS) = lead, 86.57 per cent; sulphur, 13.43 per cent. 

(Dark color.) 

Specific gravity, 6.45. 

Absorbs more sulphurous acid and 

Changes to 

Sulphate of lead (PbSOJ^lead, 68.283 per cent; ( oxygen, 10.572 per cent; 

(Light color.) ( sulphur, 21.145 per cent. 

Specific gravity, 7.13. 

Barytes (heavy spar) (BaSOJ = Ba, 65.172 per cent; ( oxygen, 22.097 per cent; 
Native sulphate of barium; ( sulphur, 12.736 per cent. 

Specific gravity, 4.7 to 4.3. 

Absorbs carbonic acid from the atmosphere, that releases the 
one atom of sulphuric acid (SO,) in its composition and 

Changes to 

Barytes carbonate (BaCO,)^ Ba, 69.10 per cent; ] ''''y^^''' 24.72 per cent; 
"^ \ a/ » 1- » ( carbon, 6.18 per cent. 

Specific gravity, 4.1. Change in volume, 5 per cent. 

Gypsum (native) ) ^ ^ ^ ^^ 20 per cent; \ ^Yf "' H'^^ ^' ^^"^' 

Sulphate of bme ) * » *- i ( gulphur, 23.52 per cent. 

Specific gravity, 2.4 to 2.08 (calcined). 

When calcined, it releases sulphuric acid and absorbs carbonic 
acid from the atmosphere and 

Changes to 
Carbonate of lime jCaO= calcium, 71.478 per cent; oxygen, moisture and car- 
Chalk or whiting ) bonic acid, 28.522 per cent. 

Specific gravity, 2.2 to 2.8. Increase in volume, 5 per cent. 



414 



CORROSIVE ELEMENTS IN OILS, 



Corrosion of Metals by Oil. 

Experiments to determine the action of oils upon copper and 
iron plates ten square feet resulted, viz. : 



AU Pure Oils. 


lO-Days' Ex- 
posure on 
Copper. 
Gain in Weight. 
Grains. 


24-Day8' Ex- 
posure on Iron. 
Gain in Weight. 
Grains. 


Almond 


0.103 

0.017 

0.010 

0.613 

0.30 

0.11 

0.22 

0.0015 

0.0485 

0.003 


0.0040 

0.08 

0.0048 

0.025 

0.005 

0.0875 

0.0062 

0.0045 

0.005 

0.046 


Colza 


Castor 


Lard 

Linseed, raw 

Neat's-foot 


Olive 


Paraffin 


Seal 


Sperm 





The result shows that the action of any oil upon any one metal is 
no guide to the degree that it will affect another metal, but that all 
metals are affected by oil to some degree. (W. H. Watson.) 

In other experiments upon commercial linseed-oils made from 
imripe seeds by the steam and dioxide-of-carbon processes, also of 
sulphuric-acid cleared oils and petroleum oils containing traces of 
sulphur, the corrosive action was from two to three times the above 
amounts on both the copper and iron. 

The corrosive effects of oils in contact with different metals at 
ordinary summer temperatures, are as follows: 



Oils. 

Mineral 

Olive 

Colza 

Tallow 

Lard 

Cottonseed . . 
Spermaceti. . 

Seal 

Whale 



Metals Not 
Attacked. 



Zinc and 



It 



n 



copper 



Bronze and tin 



ti 
{( 

it 
It 
It 



Tin 



Metals Least 
Attacked. 



Bronze 

Tin 

Iron 

Tin 

Zinc 

Lead 

Bronze 

Bronze 

Bronze 



Metals Most 
Attacked. 



Lead 
Coppei 






Tin 

Zinc 

Copper 

Lead 



CORROSIVE ELEMENTS IN SNOW WATER. 



415 



Analysis of Samples of Melted Snow from a Nxtmber of Localities, 
Showing Corrosive Ingredients. (Prof. Vivian B. Lewes, Ph.D.) 

Carbon (soot) 39.00 per cent ) 51 .30 per cent 

Hydrocarbons 12.30 " " / inert. 

Sulphuric acid 4.33 " " ) p 

Hydrochloric acid 1.33 " " [ ^?™®*^®' ^ 

Ammonia 1.37 " " ) 7.03percent. 

Metallic iron and black magnetic oxide 2.63 " " Metallic. 

Mineral matter, chiefly silica and ferric oxide. .31 .24 ** " Mineral. 

Organic matter 1 .20 " " ) 7.80 per cent 

Loss and undetermined 6 . 60 " " ) decomposable. 

100.00 " " 



Corrosive FlemerUs of Smoke and Fog, 

The composition of smoke in the cities of London and Glasgow as 
analyzed by Mr. W. R. Hutton * also includes the soot deposited by 
the smoke after being diluted by the air under the coiiditions of an 
English foggy day. 



Substance. 



Tar and oil 

Carbon 

Sand 

Iron 

Soda 

Lime 

Magnesia 

Potash 

Phosphates of lime and magnesia 

Sulphuric acid 

Chlorine 

Sulphocyanogen 

Caroonic acid 

Ammonia 

Water 



London. 



Non- 
corrosive, 



18.00% 

14.40% 
0.40% 
0.34% 
1.00% 
0.30% 
0.20% 
2.08% 
4 . 60% 

trace 
0.25% 
0.70% 
1.75% 
2.80% 



Mineral 

and 
Metallic, 
18.72% 



Corrosive 
' elements, 
10.10% 



100.00% 100.00% 



Glasgow. 



15.00% 
35.70% 

25.70% 
0.70% 

0.307o 
0.80% 

trace 
0.30% 
3 . 20% 
7.90% 
0.40% 

none 

trace 
2.80% 
7.20% 



Non- 
corrosive 
substances, 
50.70% 

Mineral 

and 

' Metallic, 

27.80% 



Corrosive 
• substances, 
21.50% 



100.00% 100.00% 



Dr. W. G. Blake found that the dust and soot in the central dis- 
trict of Edinburgh in 1902, deposited in open vessels, amounted to 
38 ounces per square foot, or about 24 pounds per year for ever}' 

♦'* Chemistry of Coal Smoke." A paper read before the Chemical Section. 
Glasgow Philosophical Society, Glasgow. Scotland. 



416 PROPERTIES OF SATURATED AIR. 

100 square feet of surface. It was highly chained with sulphuric 
add, albuminous, v^etable, and animal substances as well as soot 
and cinders. 



Tbb Weight op Air, Vapor of Water, and Mixtures of Am and Vapor at 

DlFPEHENT TEUPBaATUKES UNDER TRE ORDINARY ATMOSPHERIC PRESSDRE 

OF 29.921 Inches in the Baroubter. (Thos. Box.) 







1 

Hi 
III 


1 




"mSu« 


S( Air"sndvi^'r.' 


1 


1 


1 


.5 

II 


1 


.2 

i 


â– S:= 




















Lb>. 


32 


1.000 


.0SO7 


.181 


29.740 


.0802 


,000304 


.080504 


-00379 


263,81 


42 


1.020 


,0791 


.267 


29.654 


,0784 


000440 


.078840 


,00661 


178,18 


52 


1.041 


.0776 


.388 


29.533 


0766 


,000627 


-077227 


.00819 


122.17 


62 


1.061 


.0761 


.556 


29.365 


.0747 


,000881 


-07S581 


.01179 


84.79 


72 


1,082 


.0747 


-785 


29.136 


,0727 


,001221 


073921 


,01680 


59 54 


S2 


1.102 


,0733 


1.092 


28-829 


,0706 


,001667 


.072267 


.02361 


42 35 


92 


1.122 


.0720 


1-501 


28-420 


,06«4 


.002250 


.070717 


.032)-9 


30-40 


102 


1.143 


.0707 


2-036 


27-885 


,0659 


.002997 


.06S>^97 


.04547 


21 98 


112 


1,163 


,06W 


2,731 


27-190 


,0631 


.003946 


.067046 


.06253 


15.99 


122 


1,184 


.0682 


3,621 


26 300 


,0599 


.005142 


.065042 


,08584 


11-65 


132 


1.204 


,0671 


4,752 


25 169 


,0564 


.006639 


.063039 


.11771 


8.49 


142 


1.224 


,0660 


6,165 


23.756 


,0524 


.008473 


.060873 


.16170 


6,18 


152 


1.245 


,0649 


7,930 


21,991 


,0477 


,010716 


,068416 


-22466 


4.45 


.182 


1.306 


.0618 


16.960 


13.961 


,0288 


.020536 


,049336 


,71300 


1.402 


212 


1,367 


.0S91 


29.921 


0.000 


,0000 


.036820 


,036820 


Iiifinit« 


00,00 



OILS AND SOLVENTS. 



417 



Vehicles and Solvents. 



Substance. 



Specific 
Gravity. 



Bisulphide of carbon, CS^ 

Tetrachloride of carbon, CCl^. . . . 

Benzine, 62° B 

66° B 

Cotton-seed oil, crude 

" " refined yellow. . . 

" " water-white 

Cod-oil (tanners') 

Menhaden-oil, dark 

light 

Porgy-oil 

Poppy-seed oil 

Linseed-oil, raw, pure 

" boiled, pure 

Lucol (substitute tor linseed-oil.) 

Resin-oil, third run 

other runs 



11 
It 



ii 



<< 



Petroleum, Lima, crude 

" other brands 

Turpentine-oil, pure (C,oH,«) 

" commercial 

Water 

Other oils and fats, see page 418. 

Ammonia, 27.9 per cent 

Alcohol, pure 

" 95 per cent 

Acetic acid, hydra ted (C^H^Oj). . . . 

Sulphuric acidf (H^OSOj) 

" anhydrous (H^SO,) 

(Hvd,^hroi^c)fHClor(H,Cl.)... 

Nitric acid (HNO,) 

Carbonic acid, CO^ 



1.26 

1.56 

0.730 

0.712 

0.9224 

0.9230 

0.9288 

0.9205 

0.9292 

. 9325 

0.9332 

0.9245 

0.9299 

0.9411 

0.8993 

0.9887 

0.9910 

0.960 

0.^^39 

0.811 

0.870 

0.855 

1.000 

0.91 

0.794 

0.816 

1.10 

1.849 

1.97 

1.2 

1.52 
1 . 524 



PouDcLo per 
Gallon. 



10 

13 

6 

5 

7 
7 
7 

7 
7 
7 
7 
7 
7 
7 
7 
8 
8 
8 
7 
6 
7 
7 
8 



.513 

.00 

.09 

.94 

.696 

.701 

.749 

.686 

.741 

.781 

.786 

.723 

.759 

.853 

.50 

.2497 

.269 

.01 

.00 

.77 

.262 

.134 

.344 



7.59 
6.625 
6.808 
9.178 
15.43 
16.438 

10.020 

12.710 
12.716 



gas. 



Glycerine, (CgHgOs) Carbon, 40 per cent 

Hydrogen, 9 
Oxygen 51 

Olein 






it 



0.11636=8.594 cu. ft. 
per pound. 

I 
1.261 to 1.27 10.597 



0.93 



76 



418 



FATTY ACIDS AND SOLVENTS. 



Substance. 



Acrolein, 
Acrolic acid, 
Acetic acid, 
Margaric acid, 
Oleic acid, 
Oleic ether, 



C,H A . 
CaH.O^. . 
CjH.O,.. 
CiTHa^O,. 

CjgH^Oj. 



(Dissolves all solid fats, stearic, palmitic, 
and other fatty acids.) 

Stearic acid, CigHjjOj, 

Palmitic or > n u r^ 

Benic acid S ^le""^' 

Glycerine ether, CjH.jO; 



CjgH 



a • 



cMA- 



' 



CjyH^ or CjyHga 
85.31 per cent. 



Stearic acid, 
Stearic etlier, 
Stearine, 
Glycerine, 

Paraffin, 

_^ j Carbon, 

^ i Hydrogen, 14.44 
Fibrine: 

Vegetable, C„.aH7.,N,,.g 

Fisn, C54.7H7.2Ni5.4Sn.4. . . . 

Albumin: 

Animal, C„.,H7.iN,a.y 

Vegetable, C^.-Ji^^iN^^t 

Bisulphide of carbon, CSj 

j Carbon, 15.8 per cent. . . . 

°° ( Sulphur, 84.2 " " . . . . 

Tetrachloride of carbon, CCl^. . . . 

Linoleic acid, CioH,/).,. 

Oxylinoleic acid, CigHp^O,.. 



Specific 
Gravity. 



) 



Remarks. 



Vegetable fatty-oil 
acids. 



Melts at 112° to 149**^ 
Boils at 700° F. 



Boiling-point 109.4° to 
118.4° F. 

Boils at 170° F.; weight 
13 pounds per gallon. 



* At 32° F. 



tAt60« F. 



Proportions of Oil in Pigment Posies. 

It is often important to know the amount of oil necessary to 
form a paste with the different pigments. This amount necessarily 
varies, owing to the condition of some of the pigments before mixing 
and the fineness of grinding. The following is a general average of 
the percentage of oil in 100 pounds of commercial paste from pure 
pigments: 



White lead, pure 8 to 9 

Red lead, pure 12 

Sublimed lead 10 

Zinc oxide or white, French 16 

" " " " American IS 

Whiting paste 20 

putty IS 



per cent. 

it 



ti 



it 



tt 



It 



tt 



tt 



PROPORTION OF OIL IN PIGMENT PASTES. 419 

China clay 23 per cent. 

Terra alba 22 " " 

Barytes 8 to 10 " " 

Silica, or silex, floated 26 " " 

Lampblack G5 to 70 " " 

Drop-black 50 " " 

Gas-black 80 to 84 " " 

Mineral black 35 to 40 " " 

Graphite 30 to 35 " " 

Mineral brown 22 to 25 " " 

Vandyke brown 45 to 50 " " 

Burnt Sienna, American 35 " " 

Italian 50 " " 

Raw Sienna, American 40 " " 

" " Italian 55 " " 

Burnt umber, Turkey 42 to 45 " " 

" " American 36 " " 

Raw umber, Turkey 40 " " 

" American 36 " " 

French ochre 30 to 33 " " 

Yellow " American 28 to 30 " " 

Oxford " English 25 to 30 " " 

Indian red 20 " " 

Oxides of iron 23 to 26 " " 

Venetian red 23 to 26 " " 

Tuscan red 23 to 26 " " 

Rose pink 30 to 36 " " 

Carmine, French 50 to 64 " " 

Vermilion, American 20 to 22 " " 

English 15 to 18 " " 

" artificial (according to the specific gravity of the 

pigment) 15 to 30 " " 

Chinese or Prussian blue 50 " " 

Ultramarine blue 30 " " 

Light chrome yellow 20 " " 

Medium chrome yellow 26 

Dark or orange yellow 22 " 

Chrome green, pure (according to the shade) 26 to 35 

Chrome green, commercial (the lightest shades require the 

least oil 15 to 23 " " 

Yellow lake, French 38 " " 



tt it 

it 
tl it 



420 



ELECTRO'CHEMICAL ELEMENTS. 



Berzelius' series of electro-chemical elements and their symbols 
are as follows (those in italics are the pigment class) : 



Element. 


Symbol. 


Element. 


Symbol. 


Element . 


Symbol. 


Oxygen 


O 

s 

Se 

N 

F 

CI 

Br 

I 

P 

As 

V 

Mo 

W 

B 

C 

Sb 

Te 

Ta 

Ti 


Silicon 


Si 
H 
Au 
Os 
Ir 
Pt 
R 
Pd 
Hg 
Ag 
Cu 
Bi 
Sa 
Pb 
Cd 
Co 
Ni 
Fe 
Zn 


Manganese 

Uranium 

Cerium 


Mn 


Siuphur 


Hydrogen 

Gold 


TJ 


Silenium 


Ce 


Nitroaen 


Osmium 

Indium 


Thorium 

Zirconium 

Aluminium 

Didymium 

T lanthanum 

Yttrium 


Th 


Fluorine 


Zr 


Chlorine 


Platinum 

Rhodium 

Palladium 

Mercury 

Silver 


Al 


Bromine 


D 


Iodine 


La 


Phosphartts 

Arsenic 


Y 


Glueinum 

Magnesium 

Calcium 


G 


Vanadium 


Copper 


Mg 
Ca 


Molybdenum 


Bismuth 

Tin 


Tungsten 


Strontium 

Barium 


Sr 


Boron 


Lead 


Ba 


Carbon 


Cadmium 

CdbaU 


Lithium 


L 


Antimony 


Sodium 


Na 


Tellurium 


Nickel 


Potassium 


K 


Tantatum 


Iron 




Titanium 


Zinc 













Each metal is electro-negative to all that follow it in the list, 
and electro-positive to all that precede it, dilute sulphuric acid being 
the exciting liquid. Alloys of the metals vary the above order of 
location to a small extent, and all excitant acids also change the 
order slightly. 

The amount of electro-chemical force developed in the oxidation 
of metallic aybstances and their oxides is given on pages 353, 354, 
355. 



GENERAL INDEX. 



ACIDB 



Acrolic, 229 

Benic, 220 

Carbonic, 62, 74 

Fatty, 220. 240, 418 

Linoleic, linolenic, 219, 220, 226 

Margaric, 220 

Muriatic, 276, 277 



Nitric, 801 
Oleic, 200, 260 
Olein. 250 
Oxylinoleic. 219 
Palmitic. 220 
Pyroligneous. 100. 195 
Sulphuric, 276, 277 



Adulterants. 



Aniline, 40 

Brick-dust, 52, 58, 187 
Oment. hydraulic, 149, 164, 171 
General, 148, 811, 812 
Graphite, 138, 189 
Iron oxide. 37, 40 
Lampblack, 102 

Linseed, linseed-oil, 216, 216, 221. 226, 
236. 240, 242, 248 



Mixed paints, 96 

Red lead, 48, 57, 58 

Resinoil, cotton-seed oil, 241. 242 

Spirits turpentine, 196-198, 200 

Tin, terne-plate, 176, 183 

White lead, 65, 77-80 

Zinc oxide, 92, 94-96 



Analyses. 



Acrolic acid, acrolein. 229 
Air and smoke, 825, 826 
Asphalt, asphaltum paint, 108 
Barytes, 412, 418 
Bessemer paint, 145 

slag, 147 
Bisulphide, tetrachloride of carbon, 

203. 204. 409 
Cements, hydraulic, Portland, 149, 150 
Clays (terra alba), kaolin, 190. 409 
Coal-gas, coke-oven gas-tar, lOG-109 
Copperas, 40, 41 
Corroded iron, 888 
Euphorbium, 255 
Glycerine, 408 
Graphite, 187 
Gvpsum, 408. 412 
Inert pigments, 189. 192, 408 



Iron ores, oxides, 29, 30, 412 

Lead ores, 45, 46 

Linseed, linseed -oil, 216, 217, 218 

Litharge, 69, 409, 412 

Lucol. 417 

Manganese dioxide (pyrolusite), 227, 

412 
Mari. 190 

Mixed paints, 811-818 
Mulder's brick-dust paint, 62 
Ochre, 42 
Oil-cake, 217 
Oils, siccative, non-siccative, 217, 218, 

222, 223 
Olein, 250 

Orange mineral, 60, 409 
Petroleum, 105 
Pitch, 107 

421 



422 



GENERAL INDEX. 



Bed lead, 47, 408, 413 

Shellac, 114, 115 

Silicate of iron, 888 

Blags, blast-furnace, mineral wool, 

tetrabasic phosphate of lime, 146, 

147.409 
Snow-water, smoke, 415 
Sublimed lead, 84, 406, 412 



Tubercles, 824, 829, 878 
Turpentine-oil, dead-wood oil, 198, 

195, 196 
Umber, 48 
Vermilion, 409, 412 
White lead, 74, 75, 406, 412, 418 
Zinc oxide, 89, 90, 98, 94, 409, 412 



INDEX. 



A. 



Abbe's ref Tactometer tests for oils, 240 
Acheson's electric graphite, 141, 144, 421 
Acids, adulterants, analyses, 417 
Acrolicacid, acrolein, 229, 280, 258, 418 
Action of light on paint, 8-10 
Air and vapor mixtures, 416 
" refraction, coloring power, 8, 6, 187 
" in tunnels, 825-888 
Albumin and mucilage in oils, 218, 220, 

222, 288, 240 
Alloy corrosion, aluminum, arsenic, 

copper. 276, 882, 848, 844, 401-408 
Anchorage metal corrosion, 816, 889, 

897 
Anchorage metal electrolysis, 887 



Animal lampblacks, 103 

•• oils, 224, 246, 250, 261 
Aniline in iron oxide paint, 40 
Andrews' corrosion experiments, 881 
Angus Smith's anti-corrosive coatinir. 

128-127 
Anti-corrosive mortar, 269 
Antoxide paint, 290, 297, 800 
Argentine Republic, linseed, 218, 214 
Asphaltum and asphalt, 108-109 

paints, 104, 105, 129, 206, 

284-292, 297 
Atmospheric gases, acid, influences, 8, 

14, 92. 93, 183, 228, 265, 266, 291-296, 

8:24. 825, 897, 406, 411-418 



B. 



Barytes, blank Use. 6. 68. 77. 80. 96, 96 

140, 185, 186, 281, 282, 812, 818, 408- 

418 
Barytes-Beckton whites. 66, 67, 96, 96 
Baked Japan coatings. 119-128. 298 
Benzine, 116, 216, 814, 819, 411, 417 
Berlin blue, bone lampblacks, 99. 100 
Bessemer paint, 145-148. 296. 800 
Bisulphide, tetrachloride of carbon, 

208-207. 228, 417 
Blast-furnace, Bessemer, cement slags, 

147, 150 
Blistering and crazing of paint, 25, 216, 

819. See Peeling 
Boiling-point of coal tar, 124-127 

" turpentine, 193 
Boiling Hnseed-oil processes, 225-289 



Boiled linseed coatings, 14-16, 28-26, 
281-288. 296, 889 

Brick water-proofing, efilorescence, 118, 
164. 165 

Brickdust paints, 48, 62, 53. 187. 188 

Bridge anchorage and cable corrosion. 
389-392. 897 

Bridge anchorage and cable electroly- 
sis, 387 

Bridge anchorage and cable painting. 
816 

Bridge paints and coatings, 889-892 
" tests. 279-308 

Brooklyn gas-, water-pipe corrosion, 
392, 898 

Biirn1ngoflfmill-scaleandpaint,277.27J^ 
BuDgholeoil, 209,210,238.234,288,246 



Calcining iron ore. 80, 81, 88 
Candle-tar pitch. 129-136 
Carbonic acid in air, 325. 415 

*' in pigments, 62, 74, 411, 

418, 415, 417 
Carbonized coating. 287, 291-294, 297 
Carbon bisulphide, tetrachloride of 

carbon, 203-207, 228 



Carbonates of lead, lime, zinc, 7, 46, 

Carbon and bone-black paint, 281 

group of pigments, 98, 104- 
110 » . v^ 

litharge, red-lead mixtures, 281 
paints and varnishes, 110. Ill 
284-290 ' 

423 



<< 



424 



INDEX, 



tt 



Carter's, Clincby's quick process white 

leads, 66-69 
Cast iron not exempt from electrolysis, 

898 
Catalysis, catalytic action in paint, 

225, 267 
Caustic action of lime, cement, con- 
crete. 18, 150, 268-270 
Caustic action of roasting ore, 82-84, 

40 
Caustic compound, removing paint, 

277, 278 
Cements, Portland, etc., 149-166 
coatings, 152-168 

" porosity, 156, 156 
neat, strengtli, 154 
iron sulphide in, 151 
Cellular formation of wood, 12 
Ccrussite, cerussa, 60 
Coal-tar, analysis, tension in boiling, 

124 
Coal-tar coatings, 107-109, 124-186, 

388 
Coal-tar paints, 286 to 291 
" dipping- tank, 184 
distillation, 125-127 
Coal- and water-gas tars, 101-109, 

123-186, 400 
Colophony, 198 
Concrete qualities, 154^168 

*' voids in, produce corrosion, 

168, 159 
Cortisine, 288 
Chemical action in corrosion, 858- 

859, 862 
Chemical action in paint, 8>55, 92, 

98 
Chemical composition of glycerides, 

oik, and fats, 10, 78, 218, 237, 240, 

250, 417, 418 
Chemical changes, tests, linseed and 

other oils. See Linseed Oil. 
Chemical action of soils, 161, 878, 879, 

898 
Chinese wood-oil, 251-260 ; lacquers, 

279 
Chinese-Japanese natural varnishes, 

256-260 
Chrome blue, green, red, yellow, 8 
Chalk, carbonate of lime, 184, 282 
Chalking of white lead, 74-76, 80, 816, 

891 
Chinese, celestial, ultramarine blues, 

281, 282 
Cooper's lines in corrosion. 857 
Copal fossil resins. 111-118, 199, 224, 

242. 808, 400, 401. 405, 407 
Copperas and oxides, 7, 86. 40, 266 
Co]>per, metallic salts corrosion, 89, 

177, 264. 886, 844, 862, 879. 881, 895 



Cleaning paints, metal, 8, 12, 18, 20, 

270 277 278 
Clarifying linseed-oU, 228, 284, 285, 

248-245 
Coloring, covering jjower, 2-6, 246 
Combustion gases, effects on paint, 8, 

9, 54, 175, 294, 801, 802 
Commercial white lead, 67, 58, 65, 

77-80 
Compound paints, 8, 6. 7, 12. 80, 86, 
186, 281-288, 291. 296, 299, 811, 312- 
815, 819; marine use, 898, 899 
Compounds, use of zinc in boilers, 
method to prevent corrosion, 182, 
340, 848. 346-848 
Compounds, Wurtz's method to pre- 
vent corrosion, 188, 828. 829. 839 
Changes in cast iron from fresh- and 
sea- water exposure, 828, 824, 829, 
830, 838, 860 
Changes in paints, pigments, 56, 76, 92, 

95, 265-2u7, 411-418 
Coal, cinder, and soil corrosion, 120, 

161, 162, 378, 379, 382 
Concrete corrosion, 154-168, 889. 895 
Corroding effect of water in boilers, 

875 
Cost of water-pipe coatings, 124, 126 
" galvanizing, 177, 180 
** painting, 19, 20; by spray, 804- 

806 
" sand-blasting. 273, 274 
Crocus, 26 

Crystalline white lead, 72, 77, 88 
Cubic feet and weight of gases, 411 
Corrosion: 
Admiralty, Board of Trade, Lloyd's, 

Th wait's rules, 824, 825, 881 
alloys, aluminum, arsenic, etc., 152, 

401, 402, 405-407 
anchors and chains, 884 
anti-corrosive and anti-f ouling paints, 

400-407 
boiler sheets and rivets, 888, 884, 849 

" metal dissolved, 848 
bridge cables and anchorages, 816, 

389-892. 897, 398 
cannon. 885 
car-wheels, cold-rolled Iron, 827, 

828 
cast iron, wrought iron, steel, rate, 

822, 828. 880-832 
cast-iron cubes, 824, 825. 881 
central- station heating-pipes, 847, 

348. 874. 876 
chill-irons. 828, 829 
cold short, hot short, and laminated 

irons, 827, 882 
docks, piles, water-gates, 824-826 
Faraday's law, 868, 888. 889, 892 



<f 



tt 



tt 



INDEX. 



425 



Corrosion: 
by stress, 848-858 
floor and sidewalk beams, 161, 268- 

270, 274, 884 
French torpedo-boats, 848, 846, 406 
Irom electrolysis, 879-881 

*< sewage, 264, 828, 880 
gas- and water-pipes, 161, 261, 847 
Hambuchen's rapid method. 859-860 
hardened and burnt steel, 861, 868- 

868 
how Induced, 26-28, 156, 157, 261, 

840, 848, 860-864 
how prevented, 27, 184, 167, 170, 180. 

269, 270, 824. 828, 829, 889, 845, 

846. 879. 889. 891, 896 
of metals in contact, 262, 264, 840- 

842 
of metals in oils, 414 



Dangerous paints, 8, 110, 120, 154, 
208, 204, 806 

Dangers from electrolysis in suspen- 
sion bridges. 396, 397 

Darkening of paint, 10, 40; white lead, 
77, 78 

Davis's silicate coating. 828. 389, 896 

Dead oil in pipe coatings. 127, 128 

Dead-wood turpentine, 191 

Decay of paint. 8, 9, 13, 38, 35, 92, 93, 
120. 121, 162, 261-270, 318 

Decomposition of iron in corrosion, 
156. 157 

Delhi column, 170, 171. (Bower- 
Barff.) 

Designing paint. 6, 16, 817, 318 

Dish test for paint, 288-285 



Corrosion: 
mine pipes, 839 
New York Elevated Railway, 261, 

295-297 
Peoria water stand-pipe, 152, 153, 

874-878 
steel grillage, anchorage metal, 154. 

158-160 
stray electric currents, 886-889 
St. Lawrence River Bridge, 298, 842 

368 
Tay Bridge, 848 
rag bolts, 392 

tee-rails, cross-ties. 836-338. 886 
tin roofs, 38, 39. 48. 44, 55. 183 
tunnel-shields, metal, 157, 158, 824, 

327. 335, 389 
United States Navy Yard practice, 
I 263, 264 

D. 

, Distillation of coal-tar oils, 124, 126. 127 
Distillation of turpentine, 127-134 
Drying of linseed-oil. 131, 139, 218, 

220, 223, 235, 236, 238; under glass, 

287 
Drying of sulphuric-acid oils, 7. 280, 

281 
Driers, drying of paints, 7, 10, 43, 51, 

139, 140 
Dr. Angus Smith's anti-corrosive coat- 
ing, 124-127. 287. 394 
Dr. Dudley's boiled-oil coating. 14 
" ** paint formula, 42 

** "Wurtz's anti-corrosive process, 180, 

339 
Durable metal coatings, 287, 297, 800, 

398 



E. 



Effect of driers, dead oil, 95, 114, 229. 

235, 303 
Effect of heat on boiling, drying oil and 

paints. 8, 116, 227, 236, 287, 239. 246, 

247, 802, 803. 308 
Effect of heat and hydric-sulphide on 

paint, 54, 77, 80, 94, 116, 140, 186, 

287-291, 298. 802, 303 
Effect of sea-air, sea-water, 264. 287- 

294, 824, 384, 335. 340-344. 400-407 
Effect of sewage, condensation water, 

264, 830 
Effect of strain in corrosion, 349-352 

** *' spray painting, frost, 11, 246, 

808 
Efflorescence on brick walls, 163, 164 
Eiffel, M.. rag-bolt, 392 
Electrolytic action in paint, 81, 35, 86, 

120, 149, 155, 318, 362 



Electrolytic paints, 146, 373 

white lead, 10, 71-73 

Electrolysis, 371-399; action, 863; de- 
fined, 370 

Electrolysis in buildings, 371; bridges, 
cables. 373, 387-893 

Electrolysis, coatings, 378, 389-391, 
396. 899 

Electrolysis in concrete conduits, 395 
Faraday's law, 863, 392; 
rate of, 893 

Elcctrolvsis in gas-pipes. 384; water- 
pipe8,*87». 383, 396 

Electrolysis. Hambuchen's rapid ac- 
tion, 359-369 

Electrolysis in street tee-rails, 386-388 
" in telephone cables, 381, 

385, 888 

Electrolysis, resistance of pipe joints, 396 



426 



INDEX, 



Electrolysis, Washington Naval Ob- 
servatory. 373 
£lectro-chemicaI action for preventing 

corrosion, 180-182, 828, 839, 340, 843 
Eleetro-cbemical action of metals in 

contact, 260, 264, 340--342 
Elcctro-cbemical alloys, 401-407 
•* •* elements, 181 

Berzelius', 

420 
Electric currents, divided, stray, 880, 

886, 899 
Electric currents, jump and shunt of, 

872,888 
Electric voltages in corrosion, 873, 877, 

379, 880, 386-888, 894, 896 



Electric-furnace graphite, 142-144 
Electro - zincing processes, solutions, 

174. 176, 179. 180 
Electro-motive force in zinc batteries, 

855 
Electro-motive force in corrosion, 854 
** " ** metal under 

strain, 852, 858 
Enamel paints, 116, 121, 180, 318-321 
Engineers' indifference to cleaning 

metal 20-26, 275, 276 
Euphorbium, anti-corrosive coatings, 

254-256, 391 
Essential elements in a good paint, 1, 

20-22, 117 



F. 



Fading of paint, 10, 40 

Failure of cement coatings, 152, 154 

157, 389, 390, 392, 896 
Failure of paint coatings, 12, 20, 54-5 . 

297, 310-812, 816 
Failure of water-pipe coatings, 182 
Fatty turpentine, 199 

'• acids, oils, solvents, 10, 78. 218, 

220. 237, 288, 240, 250, 417, 418 
Feldspar, 188; refractive power. 3 
Ferredor paint (English), 290, 314; see 

also Crosley's, Armor-scale, Landers 

and Torbay's, 814, 815 
Ferric-paint test, 279-803 
Fibrin, animal and vegetable, 418 
Flake graphite, 19. 136-188; see 

Graphite 
Floor-beam corrosion, 277, 278 



tt 



Flax plant, linseed, analysis, etc., 211- 

221 
Formulae for caustic compounds, 277, 

278 
Formulae for enamels. 819, 820 

" non-corrosive cement, 268 
<< paints, 42, 295, 896 
♦* Japan driers, 208-210 
Fossil-resins, 111-118, 199, 224, 242, 808 
*• synthetical, 118 

*' varnishes, marine paint, 

818-320. 321. 400-407 
French white lead, zinc oxide, 66, 91 

•' ochre, 42. 312 
Fugitive colors, 6, 265, 266 
Fulton's flake whites, 66-69 
Furnace slags, slag cement, mineral- 
wood slag. 146, 147, 151 



G. 



Galvanizing, hot and cold processes, 
172-174. 180 

Galvanized iron paints peel, 11, 174, 204 
*• *• cost, 177; corrosion, 

174. 175. 178 

Galvanized iron, durability. 175, 181- 
188 

Galvanized iron, porosity, 178 ; protec- 
tion, 175. 176 

Galvanized iron, hardness, thickness, 
176. 178, 179 

Galvanized iron wires, loss of strength. 
178-180 

Galvanized iron, painting with bisul- 
phide of carbon, 207 

Galvanic action in cements, 154 

" " " different metals in 

contact, 830, 340, 841, 344 



Galvanic action in metal, oxide, pig- 
ments, 186. 245, 863 

Gasolene torch, burning off paint with, 
277, 278 

Gas-tank, coal-tar coating, 188, 165 

Gas coal-tar water-pipe coatings, 287- 
291, 896 

Gas- and water-pipe corrosion, elec* 
trolysis. 175. 876-886. 892, 398 

Gas-holder, spray-painting, 304 

Gilsonite, 105 

Glass, refractive power, 3 
" drying oil under, 289 

Glycerine, glycerine ether, 86, 218, 220. 
287. 288. 250. 417, 418 

Graphite, analysis, 187 ; importation. 17 

Gypsum. 6, 40. 188. 192, 269, 408. 413 

Graphite, Acheson's electric, 142-144 



INDEX, 



427 



Graphite paint, 19, 62, 136-144. 281, 

285. 291-294, 297, 406. 408 
Graphite paint tests, 140-142 
Graphite, electro-negative character, 

146 



Graphitic iron in corrosion, 829, 

388 
Graphite,8low driers, 297; yarietie8,186 



H. 



Hambuchcn's rapid corrosion, 369-369 
Heat influence on paint. 116, 227, 286, 

237. 289. 246, 247, 302, 303, 308 
Hematite ores and pigments, 30-35 
Hydrate of lead, hydrated carbonates, 

4. 5. 46. 61, 66. 71, 74, 75 
Hydraulic cement, analysis of, 149, 150 
•* coating, water-pipes, 153, 

396 
Hydraulic cracking, setting, strength, 

150-154 
Hydraulic Portland and slag cements, 

150, 151, 157. 290 
Hydraulic protecting metal, 153, 154, 

157-160, 164, 165. 395 



Hydraulic cement porosity, 155-157,396 I 261,354 



Hydraulic cement, eflfect of sulphur, 
iron sulphide in. 166. 157, 164, 171 

Hydraulic cement, Norton's experi- 
ments, 158, 159 

Hydraulic cement, marine coatings, 152 

rendering, 153, 162- 
165 

Hydraulic cement, waterproofing, 162 
*• •* use for steel grill- 

age, 153. 154, 168-160 

Hydraulic cement, use in tunnels, 159, 
160 

Hydric-sulphide effect on paint. 64, 80, 
140, 186, 287-291 

Hydrogen evolved in corrosion, 28, 89. 



Immersion test for paints, 279-284 
Inert pigments, 35, 184-192, 286, 408, 

409 
Indian red, Venetian red. 35, 37, 40, 41 
Influences that affect paints, 20-26, 

275, 276, 300-308. 411. 417, 418 
Insulating paints, 146, 373 
Inspectors and engineers, cleaning of 

metal, 20-26. 275 276 
Iodine and sapooiflcation numbers for 

oils, 243 
Iron column at Delhi, 170, 171 



Iron, changes by corrosion, 28, 324, 326, 
329, 330, 334, 8;>7. 338, 342, 392 

Iron, ores, oxides, analyses, qualities. 
26-40 

Iron-oxide pigments, condensed moist- 
ure gases, 35 

Iron-oxide pigments, durability, 6, 38 
" ** mixing defects. 37 

Iron porosity in bars, ingots, tyres, 334 
" oxide pigments. 30-38, 138, 2»1, 
282, 284. 291-293, 313-815, 406, 409, 
410, 412, 419 



Japan driers, formulae. 201-208 

• * bung- hole d riers, 209,210, 
234. 238, 246 
Japan driers, baked coatings, 119-122 



Japanese and Chinese natural var- 
nishes, 256-260 
Japanese and Chinese lacquers, 279 



K. 



Kaolin (terra alba), 190 

Kauri, fossil-resin coatings, 112, 113, 

401-404 
Kerosene, mineral oils, 105 



Knudson, A. A., electrolysis of street 

lee- rails, 386, 387 
Kirk wood, James E., coating cast-iron 

water-pipes, 123 



L. 



Labor cost in paint, 19, 21, 22 
Lacquers. 113, 256-260, 279 
Lampblacks 5, 16, 18, 49, 50, 98-102 
acids in, 98. 100 



Lampblacks, bleaching, 102; combus- 
tion by, 101; durabilitv, 102. 155, 161 

Lampblack mixtures, 49, 60, 281, 408. 
419 



428 



INDEX. 



Lampblacks, slow driers, 50, 100, 102 
Lead ores, analysis, 45, 46 ; amount 

corroded, 69 
Lead carbonate, chlorate, 45, 47, 60, 

61 66, 67, 74, 75, 232 
Lead, hydrate and protoxide, 40, 60, 66, 

71, 74, 75 
Lead pigments, 48, 74, 75, 408, 412 

adulterants, 57, 58, 65, 

77-80 
Lead, electrolytic white. 70, 72, 73 

•• "Old Dutch Process" white, 

61-75 
Lead, pulp white, 68 

" sublimed white, 83-88 

" soap, 10, 49, 74, 75, 78, 86, 235, 

237, 238. 264, 266 
Lead sulphide and sulphate, 36, 46, 

66, 94. 95 
Light refracting of pigments, 3, 4, 5 
Linoleic, linolenic, and isolinoleic acids, 

219, 220, 226, 237, 418 
Linoleate of lead. 234. 239 
Linoleum, 94, 230, 233 
Linseed, analysis of, 216, 217 ; seed and 

sources of supply, 211, 218, 215 



Linseed extraction processes. 220 ; in- 
spection, 219 ; yield of oil, 218 

Linseed fatty acids, 220 

in paint pastes, 418, 419 
" tests, 240-247 ; transparency. 
54 

Linseed, unripe seed, 222, 223, 235. 
289, 242. 314 

Linseed, sulphur effects on, 228 

Linseed-oil, 18; analysis, 220 

adulterations, 219, 240, 242, 
248, 814 

Linseed-oil, boiling data, 127-129. 220- 
246 

Linseed-oil, bung-hole boiled, 209, 210, 
233, 234, 246 

Linseed-oil, coloring power, 8 
driers. 225-286 
drying, 131, 189, 218-223. 
235-239 

Litharge analysis, 59 ; driers and 
qualities. 17, 47-49, 226. 232. 408 

Lithogen silicate paint test. 291-2^3 

Lithopone, 94, 95 186. 283, 409 

Lucol, 249-251, 297, 417 



M. 



Maize, grain oils, 79. 228 

Manganese driers, 41, 43, 206, 226, 227, 

232, 234, 238 
Manganese dioxide, analysis, 227; 

pamt, 281 
Maltha paints, failure. 110. 297 
Marine alloys and paints. 96, 400-407 
Marl analysis, 190 
Mastic resin varnishes, 117, 391 
Masonry waterproofing. 164. 165 
Meuhaaen, fish, and marine animal 

oils. 224. 242, 243. 251 
Metallic salt corrosion. 89, 152. 177. 

286 263. 264, 386. 343, 344, 362. 379, 

381, 395 



Metallic salt base of pigments. 410, 420 
Merrimac River pollution, 328 

•• Chain Bridge paint, 102 
Mineral wool. 147 
wax. 109 
** rubber paint, 289, 297 
pipe dips, 129, 180 
Mineral -oil effects tests for, 47, 57, 

241, 245. 303, 314 
Mineral-pitch oils, 104 
Mill-scale corrosion. 24, 154, 264, 271- 

278. 296. 361 
Mixed paints. 93. 810-821 
Muriatic -acid pickling, 224 
Mulder's brick-dust paints, 52, 68, 188 



N. 



Natural varnishes, 256-260 

New York elevated railway paint and 
corrosion. 262, 295-298 

Niagara Falls suspension bridge cor- 
rosion, 160, 389 



Nickel- coating, 177; alloy corrosicn, 

406 
Non-corrosive mortar, 269, 270 

water-pipe enamel, 289 



(I 



O. 



Objection to water-tests of paints, 279 
Oil and corrosion of metals, 414 
** fatty acids, and solvents, 220-224, 
417. 4l8 
Oil in paint pastes. 418, 419 



Oil coatings, 14, 15, 24, 281-288 
at the mill, 23-25 
skin experiments, 289, 

244, 300-803 
sulphur in, 245 



14 



tt 



H 



tt 



tt 



tt 



INDEX. 



429 



Oleic acid and ether, 220, 418 

Olein paint oil, 250, 251, 417 

Oxides of lead and zinc, 6, 17, 46, 74, 

8d— 97 175 
Oxides of iron. 6, 7, 10, 17, 26, 28, 80- 

87, 188, 157, 160, 281-284, 291-294, 

813-815, 884, 406, 409 



Oxidizing of paint, 7, 10 
Oxyehloride of lead, 67 
Oxygen in pigments, 411, 412 
Orange mineral, 17, 59; Paris red, 

60 
Organic matter in paints, 10 



P. 



(t 



Paints, adulterations, analyses, 7, 40, 

421 
Paints, atmospheric gases, acids, light, 

action on, 8-10. Hce A, B, Index 
Paints, characteristics. 410 

changes in. 8, 10, 55, 92, 98, 
265-268. 412, 418 
Paints, coal-tar, 287-291 

coloring and covering power, 
2-5. 186, 187 
Paints, dangerous, 8, 110. 120, 154, 

208, 204, 806 
Paints, destruction of, 8, 9, 18, 264 
decay of, 116. See D, Index 
effect of heat, frost, sea-air, sea 
water, strain, spray- painting, etc. 
See E. Index 
Paints, failures, fading, fugitive colors, 

fossil resins, rain, wind, 9, 14, 40 
Paints, formulae, 42, 295, 896 

" graphite, galvanizing. See G, 
Index 
Paints, hydraulic cement. See H, In- 
dex 
Paints, inert pigments. Sec I, Index 
labor cost, 19, 21, 22, 124, 185, 
818, 814 
Paints, mixed, 98, 810-321 

Mulder's brick-dust, 52, 53, 183 
Maltha failures, 110, 297 
mineral-oil effects, tests for, 47, 
59, 77, 241, 245, 803, 814 
Paints, mixtures, 8, 7, 18, 49, 52, 54, 55, 

80, 93, 95. 96, 187, 188, 206, 319 
Paints, peeling, crazing, livering, 11- 
14, 25, 44, 88, 92, 152, 174, 176, 216, 
227, 246, 308, 319 
Paints, removal, burning, caustic com- 
pounds, 277, 278 
Paints, statistics, 7, 17, 814 



*« 
(I 



«< 



Paints, thickness of coating, 19, 85, 

878 
Paint-tests, 16, 50, 65, 57, 68, 77, 81- 

83, 96, 279-303 
Paint tests, United States Navy Yard, 

55, 96. See T, Index 
Paint-tests, marine, 400-407 

" by water, 279-281, 284 
Painting by contractors, 811, 318 
at the mill, 23-25 
with boiled oil, 14, 15, 24, 

281-283 
Painting cement coatings, brick walls, 

164, 165 
Painting galvanized iron, 11, 174, 175 
old sign boards, 102 
by spray, 87, 304-309, 316 
Paraffin, 109, 418 

Pickling, cleaning metal, 25, 275, 276 
Pigments, inert, 35, 184-192. 408, 409 
Pipe-dips, 129-135; dipping-tank, 134, 

135 
Pitch, candle-tar, 129-186 
Pitting in boilers and water stand- 
pipes, 268. 333, 346, 360 
Polarization of turpentine, 194 
Poppy-seed and porgy oils, 80, 222-224, 

240. 417 
Porosity of iron and steel, 173, 382, 

334 
Porosity of cement, concrete, paint, 

155, 156, 173, 176, 396 
Portland cement, 150-158 
Printers' varnish, 227 
Proportion of oil in paint pastes, 8, 7, 

15. 37. 52. 54. 58, 418. 419 
Putty, 186, 186. 238 
Pyrolusite analysis, 227; paint, 281 
Pyroligneoua acid, 195-197 



Quick-process white lead or whites, 66-79 

R. 

Railway-car cleaning, painting, 277, 

278. 806 
Rape-seed, Russian seed oils, 215, 216, 

225, 245 



Red lead. 19. 22, 46, 56, 60; adultera* 

tions. 48, 57, 58 
Red-lead changes, 64, 68, 188, 808, 

412 



430 



INDEX. 



Red-lead driers, drying. 51, 52, 282 
" failures of, coating, 54-5G, 

297 
Red-lead mixtures, 54, 58, 281; lamp- 
black, zinc oxide, 49-56. 286, 299. 

312, 816 
Red lead, Mulder's brick-dust, 53 

'* proportion in paint, 58, 419 

• * * ' ready-mixed , standard ,55, 57 
Red-lead setting, 48, 49, 56, 57, 102, 

812 
Red-lead tests, 57, 281, 283, 291, 299, 

300. See T, Index 
Red lead, Paris red, orange mineral, 

60 



Refraction of light in pigments, 8-6» 

10 
Ui'fractometer, Abbe's, 240 
Resin, resin-oils, 199, 224, 242, 808, 

819, 320 
Roasting ores, pigments, 82-84, 40 
Roofing pitch and paint, 107, 129, 289, 

290 
Ruberine paints, pipe-dips, 130, 297 
Rust produces rust, 28, 157, 261, 892 

•* . mill-scale corrosion, removal. 24, 

154. 264, 271-278, 296, 861 
Rust, removal by burning, pickling, 

steaming, 275-278 
Rust, pipe- joints burst, 2t>l 



S. 



Sand-blast apparatus, 271, 272 

cleaning, 8. 13, 22, 270-278, 

296, 808, 816; costs, 278-276 
Sea- air, sea- water corrosion, 58. Bee 

£, Index 
Sea-air, galvanizing, 824-826; marine 

paints, 400-407 
Sewage, condensation water in boiler 

corrosion, 264, 328. 880. 848 
Setting of red lead, 48, 49, 50, 102 
Shellac, 114. 115, 209 
Silica floated, blanc fixe. See Barytes, 

Index 
Silica graphite paints, 188-140 

*' sand. 8, 191, 192 
Silicate coatings prevent corrosion, 828, 

839 
Slags, slag sand, slag cements, 150 
Slag paint, 290; blast- furnace slag, 147 
Snow-water, smoke, corrosive elements 

in, 415 
Soaps, lead and zinc, 10, 49, 74, 75, 78, 

86. 235. 237. 288. 264. 266 
Soapstone, 139, 191. 192 
Solvents, oils, and fatty acids, 117, 417, 

418 
Spanish-brown, 48. 44 
Specific gravities, 3, 87, 49. 59, 88, 113, 

139. 143, 188, 190, 191. 193, 203. 206, 

218, 220, 221, 224, 237, 250, 253, 824, 

326, 408-411. 416-418 
Spennrath's experiments, 300-808 
Spirits of turpentine. 116. 122. 193-202 
Spray-painting, cost, danger, durabil- 
ity, effect of cold, porosity, 87. 304- 

809, 816 



Stand-pipe corrosion, 152. 158» 874-878 
Storage of linseed-oil, 255 
Stray electrical currents, 886-888, 897 
Steam, removing and testing paint, 277 
Steel, effects of pickling, 275-277 
Strain in metal corrosion, 9, 848-858 
Sublimed lead, 88-87, 817, 412 
Suspension-bridge cable corrosion, 160 

electrolysis, 887,896, 

897 
Suspension-bridge failures, 390, 891 

'' paint coatings, 889, 

880, 891 
Substitutes for linsecd-oil, 248-260 
Sulphur compounds, 86, 87, 202 

** ** in cement, con- 

crete, 157 
Sulphur compounds in cinders, 161, 

162, 269, 834 
Sulphur compounds in oils. 245 
Sulphuric-acid oils, 2, 28, 280, 281, 244, 

245 
Sulphuric acid in cement, concrete, 151, 

156. 157, 164, 171 
Sulphuric acid in cinders, 161, 162, 269, 

834 
Sulphuric acid in locomotive exhaust, 

^ases, tunnels. 336, 337 
Sulphuric-acid effects on oils and paints, 

7. 31. 83, 50, 58. 77. 113, 206, 244, 264. 

265, 266 
Sulphuretted hydrogen, 9, 62, 140. 148, 

186. 829 
Sulphate and sulphide of lead. iron. 

zinc, 46. 62. 66, 94, 95. 276. 409.412,418 
Sulphate of lime. 3, 36, 52, 154. 186, 189 



T. 



Talc, soapstone refraction. 8 
Tables and data, 408-421 



Tay Bridge disaster, 842 
Terra alba, 190 



INDEX. 



431 



Tee-rail, cross-tie corrosion, 886-888, 
886 

Temperature and hydric sulphide ef- 
fects on paint, 54, 77, 80, 94, 140, 
186, 287-291, 298, 802. 308 

Ternc , tin-plate, and tin-roof corrosion, 
28, 88, 89, 42, 44, 55. 175, 176, 188 

Tetrachloride of carbon, drier, solv- 
ent. 206, 207, 417 

Tests for linseed- and resin-oils, 142, 
144, 199. 224, 240-247, 319, 320 

Test for paints, 10, 16, 280-286, 291. 
294, 801, 802 

Tests for white lead, 77, 81-88 

red lead. 55-58, 299, 800 
zinc oxide. 96 
turpentine, 196-198 



<• 



«< 



« t 



tt 



(< 



Thermo-clectro and chemical actions 

in paints, 31, 35, 36, 120, 149, 155, 

318, 362 
Thickness and porosity of paint, 19, 

85, 873 
Tile and brick-dust paint, 48, C2. 53, 

187, 188 
Toch's water-proof paint. 163, 165 
Torpedo-boat corrosion, 343, 846, 406 
Tunnel metal-shield corrosion, 157, 

158. 324, 327, 335, 389 
Turpentine, analysis, adulteration, 

qualities, 51, 193-201 
Turpentine, dead-wood. 195 

Douglas fir, 199, 202 
fatty, 199 



(I 



U. 



United States Astronomical Observa- 
tory, effect of stray electrical currents, 
378 

United States Navy Yard paint tests, 
96, 286, 400-407 



United States steamship, electrolysis 

in, 398. 399 
Umber. 43 
Ure, Dr., analysis of linseed-oil, 216, 

217 



V. 



Vapors, atmospheric, weight. 411, 416 
** " acids and gases, 

influence of. See A, Index 
Varnishes, qualities. 111-113, 402, 407 
" commercial brands, 116 

'* natural, non-corrosive, 256- 

260 
Vegetable oils, number of, 221-223 



Ventilation in ships, 162 ; tunnels, 887 
Vermilion, vermilionctte, 8, 60 
Viaduct corrosion, painting, 262, 263, 

296-300 
Voltages in corrosion, 372, 377, 379, 

380, 386-388, 394, 396 
Volume, weight, specific gravities, 408- 

420. See 8, Index 



W. 



Wandering electricity, 886-889 
Water and air mixtures. 416 

** corrosibility in boilers, 875 

** corrosive elements in snow and 

smoke, 415 
Water, effects on paint, 6, 8, 14. See 

E, Index 
Water- gas and coal-gas tars. 105-109, 

123-135. 400. See (J, Index 
Water- and gas-pipe coatings, 128-136, 

161. 261, 266, 347, 879, 380. 881 
Water and gas-pipe corrosion, 182, 161, 

261, 847 
Water- and gas-pipe electrolysis, 879, 

383-885, 392, 393, 396 
Water- and gas-pipe testing, 128, 124, 

132. 138 
Water in linseed-oil, 58, 128, 218, 222, 

24.5-247, 285, 818 



Water tests for paint, 279-281, 284 
Water-proofing and paint, 15, 168-165 
Weight, list and volume of gases, fats, 

fatty acids, and solvents, 220-^24, 

417, 418 
Whiting, covering power, refraction, 

3, 185 
White- lead ores. 45, 46; analysis, 74, 75 
adulterants, 57, 58, 65, 77- 

80 
White- lead carbonate, hydrate, 64, 66, 

70, 71 
White lead, ** Old Dutch Process," 61- 

66, 69, 76, 83 
White lead, electrolytic, 70-73; sub- 
limed, 83-88. 317 
White-lead tests, 77, 81-88 

qualities of a good, 74-78 
in marine paint, 96, 289 



<< 



X 



432 



INDEX, 



White- lead, zinc oxide, and barytes 
mixtures, 7, 77, 80, Oa, 06, 812, 816 



Wooden-structure paints, 6-8, 84, 86, 
80, 81, 96, 246 



Y. 



Yellow pine, area available for turpen- 
tine, 198 
Yellow pine distUlation, 195 



Yellow pine, painting, 80, 81, 96, 246 

'• paint size, 80, 81 
Yellowing of oil and white lead, 7, 10, 40 



Z. 



t< 



<< 



(( 



2^nzibar, fossil resins, 111-118 
Zinc, metallic 89; consumption, 91 
ores, calamite, 89; zincite, 91 
alloy in galvanizing, 177 
carbonate, 89; sulphide and sul- 
phate, 77, 92, 9.S, 95, 96, 409, 412 
Zinc, electro-chemical and galvanic 
action in, 98, 174-183; E. M. F. in 
batteries, 853, 355 
Zinc, galvanizing, 172-188 
' ' oxide, history, 91 ; French brands, 
91,92 
Zinc-oxide adulterants, 92-96 

changes in, 77, 79, 80, 93, 
413 
Zinc-oxide pigments, 77, 78, 80-07 



Zinc-oxide marine paints, 96, 286, 405- 

407 
Zinc oxide and white lead, 58, 80, 81, 

92-96, 312, 816 
Zinc-oxide tests, 96, 97, 405, 407 
Zinc, use in steam-boilers, 128, 839-846 

•' *' to prevent corrosion, 173-181, 

844-848 
Zinc-salt paint driers 282 

" sheets and roofing durability, 28, 

181-183 
Zinc soaps. See L and 8, Index 

" whites and lithopone, 94, 95, 186, 

2r>2. 409 
Zincing solutions, 176 



ADVERTISEMENTS. 



INDEX TO ADVERTISEMENTS. 



PAGE 

Berry Brothers, Limited 2 

Detroit Graphite Mfg. Co i 

Eagle White Lead Company 8 

International Acheson Graphite Company 7 

Lowe Brothers Company, The 9 

Michigan Paint Company 4 

Picher Lead Company 6 

RiNALD Brothers 3 

TocH Brothers 5 



Superior 
Graphite 
Paint 



THIS is the well-known L. S. G. brand. Its basis 
is a natural ore, the peculiar ingredients of which 
best adapt it to the making of a durable paint. 

It does not crack or peel, is unaffected by heat, 
moisture, sulphur fumes, acids, alkalies, drippings, etc. 

It should always be specified as shop and field coats 
for the structural steel of buildings and bridges. 

It is unequalled on corrugated iron, steel cars, tanks, 
pipes, roofs, refrigerator plants, elevators, storage bins, etc. 

Among the many structures upon which SUPERIOR 
GRAPHITE PAINT has been recently used are the Flat- 
iron and Macy buildings of New York, the Government 
Printing Office and Stoneleigh Court, Washington, D. C., 
and the plant of the Niagara Falls Power Co. 

We have an Interesting booklet which should be on 
the desk of every engineer, architect and master painter. 



S!^'* DETROIT GRAPHITE MFG. CO. eSr" 

*"*" DETROIT. MICH. ^ ^" 



We would like to hear from consumers of 

Japan Driers 

who want the best goods procurable. 

An experience of nearly 50 years in 
Varnish and Japan making has taught us 
many things that only time can impart. 
** Practice makes perfect " is one of the truest 
of the old saws. 

We offer reliable Japan Driers of 
all grades and of uniform quality. 



BERRY BROTHERS, Limited 

NEW YORK BOSTON PHILADELPHIA BALTIMORE 

CHICAGO CINCINNATI SAN FRANCISCO ST. LOUIS 

Factory and Main Office, 
DETROIT. 



Manufacturers of every grade of Varnish and Japan 

for every use known* 

2 



SSessemer iPa/ni 

(REQISTERBD TRADE-MARK) 

Has been in acttsal use on bridges and iron and steel budd- 
ings for over ten years and still retains its elasticity and 
protecting qualities^ where Graphite, Red Lead, etc«, have failed 
Our pamphlet, ^^Data on Bessemer Paint,^^ is yours for 
the asking* 

iPorceiain Snamei iPaint 

"THE ONLY ASEPTIC ENAMEL MADE." 

A Liquid Porcelain made in white and colors, which dries 
of its own accord into a smooth porcelain surface* Endorsed 
and used by the leading hospitals and numerous house owners, 
hotel proprietors, etc*, in kitchen, bathrooms, etc* Send for 
our pamphlet, ^^AH About Porcelain Enamel Paint^^ 

Concentrated Saivamc ^Primer 

Prevents peeling and blistering of paint applied to galvanized 
iron* 

Uechnecal iPa/nts 

Are our specialty* We do not make **A Line of Paints,^ like 
other manufacturers, but only ^^ Special Paints for Special 
Purposes*** 

tW If you have any difficult problems to solve in the paint 
line we shall be glad to place our experience at your command 

RINALD BROTHERS 

JJ42-n4^ N. HANGOCK ST., PHILADELPHIA 

3 



ft 



"MADE IN FLINT 

Our PaJnt protects metOLl •••••• 

We protect the user with guarantee 



MIGHI6AN 6RAPHITE PAINT 



MIXED and Pulverized at L'Anse, Baraga County, 
Michigan, on ttie stiores of Lulce Superior. 

GROUND ttirougli stone paint-milla at Flint, Genesee 
County, Michigan. 

THINNED with absolutely pure linseed oil. 

SOLD either as paste or ready for application. 

PRtCE lowest consistent with care of manufacture and 
the use of absolutely pure oil (oil quality is as 
important as the pigment used). 

Liquid, per gallon, $1.05 

Paste, '* pound, .08 

SPECtPICATiONS.—To insure use of this paint specifica- 
tions should read, in addition to firm name, 
"Made in Fiint." Barrels found on Jobs not so 
stenciled do not contain our goods. 

MICUieAN PAINT COMPANY 

aNCORPORATED) 

FLINT. MICHIGAN 



We make a specialty of grinding and preparing paints on 
special formulas furnished by engineers or other customers. 

WB WANT BUSINESS. 

4 



R.I.W. Damp Resisting Paint 



FOR. INSIDE 

AND 



KoNKERiT Coating 

(tradk mark rco.) 

FOR OUTSIDE 



ON WALLS 

(Damp Proof) 



IRON WORK 

(Rust Proof) 



LIMESTONE 

(Stain Proof) 



INSULATION 

(Elec. and Cold Storage) 



OF YOUR. BUILDING 



The Modern Paint for Modern Buildings 



TOCH BROTHERS 



PAINT MAKERS SINCE 1S48 



£^Q 




• r% 



, 470 & 472 West Brootdwoty 



NEW YORK 

5 



Perpet\ial 
Protection for 
Iron e.nd Steel 




PICKER'S NATURAL BLUE LEAD (steei coior) 

is a chemically stable product. It protects the 
oil— imparts to it permanent life. 

PICKER'S NATURAL BLUE LEAD can be 

ground in paste form and kept indefinitely 
without hardening. Pound for pound it will 
coyer 20% more surface than any other lead 
pigment, and with boiled linseed oil it makes an 
indestructible paint for iron arid steel surfaces. 

It can be obtained from the manufacturers direct or 
from any reputable paint grinder — in dry, paste 
or liquid form. 



A BOOKLET OF DETAILED INFORMATION FOR THE ASKING 



RICHER LEAD COMPANY 

EASTERN SALES OFFICEi WESTERN SALES OFFICE: 

100 WILLIAM STREET TACOMA BUILDING 

NEW YORK CHICAGO 

Works: JOPLIN, MO. 

6 



flCftESON GRAPHITE 



Fellow Reauers; 

We have but one suggestion to add to the pages of this 
book — Determine a Standard for Graphite Paints. There is 
to-day absolutely no standard. The so-called graphite pigments 
contain anywhere from 20% to 80% of impurities and adulter- 
ants of a heterogeneous character. Sometimes they contain no 
graphite. In consequence many engineers are imposed upon, 
and there are many failures of such protective coatings. 

We are naturally interested in this matter, as Acheson 
Graphite Paint Pigment fully meets all the conditions which have 
ever been found of value in graphite pigments. It Is manufactured 
in the Electric Furnace and contains all the merit, with none of 
the disadvantages, of the natural graphites. We guarantee the 
Purity, Uniformity, and Inert Quality of Acheson Graphite. 

Write us for samples for test, prices, or any further in- 
formation desired. 

INTERNATIONAL ACHESON GRAPHITE COMPANY 

NIAGARA FALLS, N. V., U. S. A. 



I 



NDEPENDENT OF ALL COMBINATIONS 




"^^/TY 15,000 



£AGLi: 

WHITE LEAD 

COMPANY 



1008 to 1030 Broetdway CincinneLti. O. 



Corroders by the ^^Old Dutch Process" 

PURE WHITE LEAD, DRY AND IN OIL 
R.ED LEAD, LITHARGE, AND ORANGE MINERAL 



OFFICES AND WAREHOUSES: 

NEW YORK CITY 54 MAIDEN LANE 

AUSTIN REMSEN, Manager. 

PHILADELPHIA 14a N, FOURTH STREET 

T. E. BANNAN, Manager. 

BALTIMORE 447 NORTH STREET 

QEO. O. 5HIYER5, Manager. 

BUFFALO 16 BUILDERS' EXCHANGE 

A. S. QOLTZ, Manager. 

PITTSBURG, 

PITTSBURG PAINT SUPPLY CO., Agents. 

CLEVELAND, 

A. T. OSBORN CO., Agents. 

CHICAGO 135-137 NORTH PEORIA STREET 

E. B. BENNETT, Manager. 

ST. LOUIS, 706 N. ELEVENTH STREET 

F. L. POWERS, Manager. 

KANSAS CITY ioia-14 WALNUT STREET 

W. R. MCDONALD, Agent. 

NEW ORLEANS 308-310 GRAVIER STREET 

JNO. R. TODD Sl BRO., Agents. 

8 



LOWE BROTHERS PREPARED 
LINSEED OIL PAINTS 

FOR THE PRESERVATION AND PROTECTION OF 
STEEL IN BUILDINGS, BRIDGES, RAILWAY 
EQUIPMENT AND GENERAL MANUFACTURING 

fHE ACKNOWLEDGED "HIGH STANDARD" 

Among these produdi are Includedi 
RED LEAD METAL PRESEKVA- 

TIVE (two colon). 
BLACK METAL COATING 

No. 1407. 
GRAPHITE PAINTS. 
CARBON PAINTS (fburcolon). 
OXIDE OF IRON PAINTS (three 

colon). 



All are Linseed Oil PainU with selected pig- 
ments only. They are made from legitimate ma- 
terials and based upon the theory that the solids 
are coefficient with the liquids in producing the 
best results, and that their quality is as depend- 
ent upon their physical as upon their chemical 
structure. 

The principles upon which they are made and 
used are fully discussed In "Hints on Painting 
Structural Steel", "Suf^estlons for Specifications** 
etc. Sent on application. 



THE LOWE 
BROTHERS 
COM PANT 

DAYTON. O, 
NEW YOIIK 
CHI C-\ G O 
KAN8A8 CITV