198 



NATURE 



[M. 



b«en und'-rtaken with the view of settling thcM: impurtaiu 



f K>' n r - . 



IS. K, l". Law, W. H. Mcrrct, and \V. 1'. Digby 

 - ^lutlic•d welds, and in their paper present the resuhs 

 ot their investigationii. Defining a weld as the actu^.! 

 fusion together of similar or allied inetal«, the authors have 

 carried their investigation!! into the region of the ch'iracter- 

 i«tic microstructure of both good welds and u'-.AUcresbful 

 attempts to produce welds, a subject o«i whl.h there has 

 hitlii-rto hi'A'W very little work done. No matter what the 

 process may be by which two metals are welded together, 

 there must always be an area, more or less sharply defined, 

 of altered molecular tx>nstruction. The authors show that 



• iich process of welding has its own halUmark. For 



• xainple, it is possible to say whether an electric weld has 

 b<tn made by an arc or by a resistance meth«xl from the 

 ' A-idence afforded by polishing and etching alone. Without 

 annealing to restore the original structure, acetyfene and 

 water-gas welds have each their own marked character- 

 istics. Resistance welds are seemingly less prone (with 

 the exception, perhaps, of acetylene welds) to oxidation, 

 but the extrusion of the metal renders gooti working, while 

 the metal is still plastic, of supreme importance. Arc 

 welds are most prone to oxidation, and many will hesitate 

 to rely on such a process in those positions where corrosion 

 is likely to occur. When the welded metal is not likely to 

 be subjected to corrosion, the excellent fusion of the metal 

 renders the method commendable. Flame welds should 

 receive adequate working and manipulation while in their 

 heated rondititin. Water-gas welds may be abnormal 

 through the use of oxidising flames; acetylene welds 

 <rrtainly require annealing to break down the crystalline 

 structure in the vicinity of the weld. 



Prof. 11. ('. n. Carpcnte'- has continued^is investiga- 

 tions on the growth of cast irons after repeated heatings. 

 His principal results up to date may be summarised as 

 follows : — phosphorus tends to diminish growth ; sulphur 

 is never present in commercial cast irons in sufficient 

 quantity to have more than a small influence on growth, 

 which is, however, in the direction of retardation; man- 

 ganese always retards the rate of growth, and diminishes 

 the absolute amount in the majority of cases. Dissolved 

 gases have no influence on the growth of an iron containing 

 more than 3 per cent, of silicon ; if the silicon does not 

 exceed i per cent, they may be responsible for a growth 

 of at least 10 per cent. The simplest and most rapid test 

 for forming an opinion as to the growth that is liable to 

 take place in any particular grey iron is to estimate the 

 silicon, and then read off the approximate growth from 

 the following table : — 



If the iron contains 03 per cent, of phosphorus and 

 upwards, the growth will be from 2-5 to 4 per cent, lower 

 Than the above figures, and if more than 05 per cent, of 

 manganese is present, the rate of growth wilfbe diminished 

 and the amount of growth somewhat lessened. An alloy 

 containing 2-66 per cent, of carbon, 0-587 per cent, o'f 

 silicon, and 1-64 per cent, of manganese, showed no signs 

 of growth after 150 beats, but, on the contrarv, a slight 

 contraction, viz. about 013 per cent. It is a tough 

 material, and its mechanical properties were improved bv 

 this treatment. It begins to freeze at about 1346° C. and 

 appears to be a suitable material for annealing ovens, 

 rolls, fire-bars, and the grids of muffle furnaces. Probablv 

 It could be used for ingot mouUK in ;.n iron foundry with- 

 out cracking. 



The influence of impurities on the corrosion of iron is 



dealt with in a paper by Mr. J. W. Cobb. Interpreting 



the results of the author's experiments on the basis of the 



electrolytic theory of corrosion, it may be stated that pure 



NO. 2168. VOL. 861 



iron in 'definitely elect ro-po«itive tu must of its impurii 



tried (excepting ferro-inanganewj were also electro 

 to pure iron. With the sul{>l>ide and silicate of \\\ 

 little or no current flowtid, M and «o ) 



ferro-manganese wcrre found electro-pi. 



iron. Kvery piece of cor ' ' 



I- fleets with any other, ai 



the same piece were alw. . 



when the other conditions were satisfied. 1 1 

 an impurity determines so many corrosion c» 1 

 and .so its influence depends more on quality and 

 tion than on quantity; thus a more homog<-n<-ous ii 

 if chemically less pure, may be more highly r«-- 

 corrosion. Other pap«;rs Ixiaring on corrosion \\ 

 tributed by .Mr. I'. Longmuir and by .Messrs. J. 

 Friend and J. H. Brown. 



•Mr. W, H. Matfield gives experimental resyli- 

 influence of vanadium upon the physical properties, of • . 

 iron, .^s an instance where vanadium has increa«*<l t 

 life of locomotive cylinders, a «. i-d where • 



made of cast iron not treated v um wore 1 



per Kjo.ooo miles, whereas \ last-iron 



showed only microscopic wear after running 200,0. 

 The present experiments show that additions of \ 

 have a definite influence upon the physical pro) 

 cast iron, and that this influence is mainly that . 

 ing the carbon to p«'rsist in the combined state. I 

 persisting carbides, physically, do not differ materi; 

 from the normal carbides found in the cast iron ; owi: 

 however, to tiie actual presence of much of the vanadii 

 in the carbide, that carbide is rendered more stable. 



.Messrs. \. .McWilliam and E. J. Barnes give records o( 

 a lengthy series of experiments on the influence of 0-2 per 

 cent, vanadium on steels of varying carbon content 



.\ paper on the chemical and mechanical relations ol 

 iron, chromium, and carbon is contributed by Profs. J. O 

 .Arnold and .\. .\. Read. This pap-r is in continuation o 

 the work of the authors already published, and gives an 

 account of a number of experiments to determine the com- 

 position of the carbides separated from a series of annealed 

 steels containing various percentages of chromium, the per 

 centage of carbon being practically the same in each. Th< 

 mechanical properties of these alloys under static an< 

 alternating stress, and their microscopic features, have als< 

 been investigated. 



Iron-silicon-carbon alloys are dealt with in a paper b^ 

 Dr. W. Gontermann. Some of the work performed at th< 

 Institute of Physical Chemistry at Gotlingen has alread] 

 been published, and the present report contains furthei 

 particulars. The paper contains many diagrams and phol 

 graphs of models showing graphically the properties of thi 

 series of alloys. 



The magnetic properties of some nickel steels, and not" 

 on the structures of meteoric iron, form the subject of ; 

 paper by Messrs. E. Colver-Glauert and .S. H-* 

 Berlin. A 5 per cent, nickel steel is hardest (:; 

 ally) when quenched in the neighbourhood of «)i'»>"' t 

 Quenching from higher temperatures results in a softer 

 material. The changes which occur during thermal treat- 

 ment of a 25 per cent, nickel-iron alloy are of a far mor- 

 complicated nature than has been thought previous!} . 

 .\t high temperatures there probably exists a product 

 which may be preserved by rapid quenching, and is then 

 strongly inagnetic, and persists to the temperature of liquid 

 air. This product does not exist in the region betweer. 

 about rx>o° C. and qoo° C. There is very 

 tion between the magnetic properties and n 

 structure. There is no sharp magnetic ch;i_, 

 this alloy below zero, but the permeability gradually in- 



1 



creases as the temperature decreases from about —50° C- 

 to — 180° C. The magnetic properties of a 33 per cent. 

 nickel-iron alloy are only very slightly afTected by thermal 

 treatment. The microstructures of commercial nickel 

 steels are practically the same as those of meteoric iron. 



Messrs. A. McWilham and E. T. Barnes . .nnlrK tlv ': 

 series of papers on steel with another on th 

 heat-treated 3 per cent, nickel steel. 



Messrs. F. A. Daubing and E. V. Roy. of 

 France, give an account of a process for the d 



