March igth, 1918.] Proceedings. xi 



and form channels by which the corroding liquid or gas penetrates 

 to the interior, and this is intensified by alternate expansion and 

 contraction due to heating and cooling, and by vibration. This 

 increase is most pronounced where the continuity of structure and 

 cohesion is least — i.e. at the graphite flakes. It was shown by 

 photo-micrographs and actual specimens of corroded material, that 

 the concentration occurs, and that corrosion follows these segrega- 

 tions and the graphite. Examples in which the graphite plate 

 occupied the middle of the corrosion were pointed out. 



Specific instances in which the failure of cast-iron vessels was 

 due to the increase in volume resulting from the corrosion, and the 

 influence of the structure due to the segregation and coarse graphite, 

 were dealt with and specimens shown. Analyses and examinations 

 showed that the collapse of the vessels was due to these causes. 



Attention was also directed to the high silicon iron now used 

 for chemical plant, and segregation was shown to take place to a 

 marked extent. The author showed that the failure in many cases 

 investigated was due to the presence of graphite and phosphide. 

 Separated pellets of phosphide taken from cavities in metal con- 

 taining 13.6 silicon and 0.41 phosphorus contained over 4.1 per 

 cent, of phosphorus and only 10.45 of silicon. These were attacked 

 while the silicon iron itself is but little affected. In a series of tests 

 it was shown that, in the same metal, the amount of phosphorus 

 removed by the corroding liquids employed was many times greater 

 than the proportion in the mass of the metal, thus showing that the 

 cavities and concentrations formed the line of attack, which led to 

 the ultimate failure of the metal. 



Ordinary Meeting, March 19th, 1918. 



The President, Mr. William Thomson, F.R.S.E., F.C.S., F.I.C., 



in the Chair. 



The President referred to the death, on March 12th, of Mr. 

 George Stephen Woolley, who was elected a member of the Society 

 in i860. 



Mr. Thomson exhibited two specimens of .what were repre- 

 sented as minerals which had been washed down from a mountain 

 in Angola (a province in West Africa, south of the Congo). The one 

 was a cube of Iron Pyrites about a cubic inch covered with a hard 

 layer of the proto and per oxides of iron about a quarter of an inch 

 in thickness. The pyrites was free from arsenic. 



The other was a specimen of what I found to be metallic iron 

 associated with silicium and graphite, but containing no nickel, 

 and which, it was suggested, might have been manufactured by 

 the natives, and not native iron, This might possibly be so, but 



