1843.] 



THE CIVIL ENGINEER AND ARCHITECT'S JOURNAL. 



387 



tervals of professional avocations. The experiments, originally undertaken at 

 the request of the British Association, had reference merely to the action of 

 air and water on iron, in various states and under several conditions ; but in 

 the progress of research, the author has extended the investigation to many 

 other connected branches of practical importance, such as the several modes, 

 or assumed modes, of protecting iron from corrosion, in which he has pro- 

 posed improved methods dependent upon known principles, which are deve- 

 loped at length, and which appear likely to be of great importance, in rela- 

 tion to the circumstances affecting the durability and corrosion, whether 

 general or local, of iron ships, to which he has devoted much attention. 

 The protective powers of various paints and varnishes have also been deter- 

 mined in many conditions, and the peculiar circumstances of corrosion pre- 

 sumed to attach to railways bars, are partially investigated. 



The main numerical results of these investigations are given in two vo- 

 lumes of Tables, which show on inspection the absolute and relative losses 

 by corrosion, in given times, and under six several conditions of experiments, 

 on nearly all the various qualities of cast and wrought iron, and steel, made 

 at the most important works in Great Britain. During the experiments, the 

 metals were exposed at two several periods of 387 days, and of 732 days 

 respectively, to the action of — 



1. Clear sea water, at temp. 56° to 58° Fahrenheit. 



2. Foul sea water „ ditto. 



3. Clear sea water ,, 110 to 125 ditto. 



4. Clear river water „ 32 to 68 ditto. 



5. Foul river water „ 36 to 61 ditto. 



6. Freely exposed to the atmosphere, and its precipitations, at 



Dublin. 



Tables are also given containing the amounts of corrosion of cast and 

 wrought iron, in sea and fresh water. 



1. When coated with zinc, or galvanically protected by its contact. 



2. When painted and varnished in various ways. 



3. When in contact with various definite alloys of copper and zinc, and 



of copper and tin, as in brass and gun metal, which are both elec- 

 tro-negative to iron in water. 



4. Of cast iron, with the surface variously modified by the method of 



coating, as by chilling, &c, or with the surface or coat removed by 

 planing. 



5. Of the specific gravities, rigidly ascertained, of all the specimens of 



cast and wrought iron and steel, experimented on, and determina- 

 tions, for the first time made, of the changes in density, produced 

 by casting iron under a variable head of metal, and of the effects 

 on density of changes in the mass or bulk of castings in iron ; all 

 which, the author has shown, affect the ratio of corrosion of a 

 given sort of cast iron. 



These tables, which the author intends to contain the chief information 

 requisite for the engineer, to enable him in practice to allow for the loss of 

 his structures by corrosion in any given time, and to choose the most desi- 

 rable irons, &c, are followed by others, which condense into one view the 

 whole results arrived at. and fit them for practical reference. 2 



Another principal object held in view in the tabulation of these results, 

 and effected by the author, was by discussion of their contents to discover 

 upon what variations in the texture, density or chemical constitutions of the 

 metal, maximum and minimum corrodibility under given conditions de- 

 pended ; for this purpose numerous exact analyses of the cast and wrought 

 iron, &e„ of maxima and minima corrosion, became requisite. These have 

 been made by the author, and are given in separate tables, together with the 

 details of the methods adopted for obtaining correct results ; a matter of 

 admitted difficulty in the case of the analysis of iron. 3 These analyses show 



- These tables are not susceptible of abstract, but they are being prepared 

 for publication in the forthcoming volume of the Transactions of the Insti- 

 tution of Civil Engineers. 



3 The author thus describes the mode of analysis practised by him : — 

 " The method adopted by me, in most cases, was a modification of Reg- 

 nault's process, which consisted in mixing the east iron, finely pulverized, 

 with about twelve times its weight of chromate of lead properly prepared, 

 and mixed with a little chlorate of potash. This is burnt in an ordinary 

 combustion tube, in the remote extremity of which some dry powdered chlo- 

 rate of potash is placed, and heated after the combustion has been completed, 

 so as to pass a current of oxygen over the ignited mass. This precaution is 

 indispensable with the harder and denser irons, containing most of their 

 carbon in combination. The total amount of constituent carbon is thus ob- 

 tained, and weighed as carbonic acid: but this consists of graphite and of 

 combineil carbon. By a separate assay, the graphite is obtained, on solution 

 of a weighed portion of the metal in nitric acid, as residue, consisting of 

 graphite, extractive matter (from the carbon of combination) and silica, and 

 occasionally some oxydes of combined metals. The residue is filtered and 

 washed, boiled in caustic potash, by which the silex and extraction are taken 

 dp. The graphite remains: it is ac;ain washed with dilute muriatic acid, 

 theu with water, and weighed after drying. The difference between this and 

 ihe total amount of carbon given by the combustion is equal to the carbon 

 of combination. 



" For the other constituents, after a preliminary qualitative trial, about 

 120 grains of the cast iron were dissolved in nitric acid, evaporated to 

 dryness with a strong heat, and ignited in a platiua crucible with three and 



that corrodibility does not depend upon the proportion of constituent carbon 

 in cast iron, and still less upon that of the ether foreign matters usually 

 found in it; but upon the stale in which carbon exists in the compound; 

 upon the state of aggregation of the whole mass; upon the density, and 

 upon the voltaic uniformity, or otherwise, of the surfaces exposed to corro- 

 sion. Thus the same sort of cast iron corrodes much faster, in given condi- 

 tions, if cooled irregularly, and faster than it does when cooled uniformly 

 and slowly. 



Hot or cold blast produces very little difference iu corrodibility of cast 

 iron, and this results chiefly from difference in density; recollecting that 

 carbon exists in cast iron iu two very different states, viz., as diffused gra- 

 phite in a crystallized form, and as combined carbon ; that the dark grey 

 and softer irons contain more of the former, while the harder and brighter 

 irons have more of the latter; that the latter kind have much less uniformity 

 of surface, when cast under similar conditions, than the former; while the 

 highly graphitic irons, though more uniform in large specimens, are the least 

 dense and softest in texture — we arrive hence at the ultimate choice that 

 the bright grey irons of high commercial value, while they are in all, other 

 respects the most useful for construction, are also the most durable when 

 exposed to the action of air and water. The second prolonged period of 

 immersion of all the specimens was necessary, in order to determine the "law 

 of progression of corrosion, with respect to time." The author finds that 

 where the coat of oxyde and of carbonaceous matter or plumbago formed, 

 is constantly removed from the surface of cast iron exposed to corrosion in 

 air and water, the progression of the latter is a decreasing one, because as 

 the metal is removed, the inner portions become more uniform in texture, 

 and fewer minute voltaic couples are formed ; but where the oxydes and 

 plumbago remain untouched, these being both electro-negative to the metal, 

 nearly equilibriate the effect of the regular texture ; and thus the rate of 

 corrosion remains uniform, or is nearly in direct proportion to the time of 

 reaction. This is demonstrated experimentally, and is most forcibly exhi- 

 bited in corrosion by sea water. Hence in practice, cast iron immersed 

 without any protection, will corrode less if occasionally scraped and cleaned; 

 or if in a tide- way, than if untouched and in still sea water. 



The rate of corrosion, as dependent on the metal itself, is a minimum 

 when the cast iron is most uniform and hard, and free from suspended gra- 

 phite, and as dependent upon the water in which it is immersed ; is a maxi- 

 mum in foul sea water, and a minimum in clear river water, both being at 

 mean temperature, and containing nearly the same volumes of combined air 

 and carbonic acid. The kyanized oak boxes, 2 in. in thickness, in which the 

 specimens were immersed in Kingston Harbour, were eaten through in about 

 two years by the limnoria terebrans. Cast iron freely exposed to the weather 

 at Dublin, and to all its atmospheric precipitations, was corroded nearly as 

 fast as if in clear sea water, when the specimens in both cases were wholly 

 unprotected. 



The results of experiments on wrought iron and steel, show that they con- 

 sist of two or more different chemical compounds, coherent and interlaced, of 

 which one is electro-negative to the other. The electro-positive body being 

 that which suffers first from corrosion, the electro-negative portions of the 

 iron and steel remain bright, and hold a perfect metallic lustre until the 

 whole of the other portions are removed, or at least are so to a great depth, 

 when they begin likewise to oxydate. In general the finer the quality of 

 wrought iron, and the more perfectly uniform its texture, the slower and 

 more uniform is its corrosion in water; minute differences in chemical con- 

 stitution produce little change in this respect. Highly silicious wrought 

 iron, however, corrodes very locally, and appears to be partially defended by 

 a thin coat of silex formed on it. Fagoted scrap bars, made from best Staf- 

 fordshire rivet iron, was found of all the irons experimented upon, to be the 

 most durable ; next to this was Low Moor boiler plate, and it is thence pre- 

 ferable for iron ship building. Foul sea water, evolving sulphuretted hy- 

 drogen, gives the maximum corrosion of wrought iron and steel. The con- 

 tact of soft putrifying mud appears to be still more destructive. Steel gene- 

 rally corrodes more uniformly and slowly than wrought iron. Hardened 



a-half times its weight of carbonate of soda. After cooling, water is poured 

 over it, which carries off the excess of alkali and an alkaline phosphate or 

 sulphate, if the iron contained sulphur, which should lie ascertained before- 

 hand, leaving the peroxide of iron to be separated by filtration. The filtered 

 liquor must now be bailed for some time to destroy the manganesiate of 

 potash in solution, and precipitate the manganese; again filtered, nitric acid 

 added, evaporated to dryness, and silicic acid separated, if any exist, on 

 heating with water, after moistening with acid in the usual way. Ammonia 

 is now cautiously added, and if the iron contained aluminum, a basic phos- 

 phate of alumina precipitates. The solution, again filtered, is acidulated 

 with acetic, and the phosphoric acid precipitated by acetate of leal. From 

 the phosphate the phosphoric acid cannot be estimated with certainty ; it 

 was therefore converted into sulphate of lead, and the phosphoric acid got 

 from its weight. 



" The silex and manganese were always obtained by precipitation from Ihe 

 iron, 8k„ in separate assays; the method with benzoate. or succinate of 

 ammonia, though inconvenient, is one of the best where the amount of iron 

 is considerable. Leibig's process of separation, by boiling with carbonate of 

 baryles, succeeds very well, and presents no difficulties; but whin the 

 amount of manganese is so very small in proportion to the iron, I preferred 

 the former mode. The iron itself, from its inconvenient bulk, was generally 

 estimated from the other constituents. SeparaJe assays are also best made 

 for sulphur or earthy bases; but as far as my observation goes, these are 

 extremely rare in British cast-iron. 



