June 2, 1892] 



J>JA TURE 



11=; 



figures as showing the best results, and to serve as a guide ; it is 

 doubtless unnecessary to say that they are not conclusive stand- 

 ing alone. The riveting tests given are valuable, but these are 

 of a nature which cannot be epitomized. 



Colonel Dyer's paper might well have been longer, as the 

 subject of it is one of considerable importance. Pure iron is a 

 substance at any rate difficult to get. Sir Lowthian Bell has 

 said he has never met with absolutely pure iron. Commercially 

 pure iron, or what might be called practically pure iron, is not 

 uncommon. Colonel Dyer's object was to obtain a pure iron in 

 order to determine the value of alloyi=. By working on the lines 

 which he had followed, the author hoped that pure iron and 

 steelmay.be produced at reasonable cost. In the first experi- 

 ments the furnace was charged in the ordinary manner with pig 

 and scrap of fairly good quality, and the charge was worked 

 slowly, care being taken to keep the slag well saturated with 

 lime by liberal additions of limestone. The phosphorus was 

 reduced during the process, but the result left much to be desired 

 in other respects. Charges composed of from one-half to four- 

 fifths of good scrap, and one-half to one-fifth of good Swedish 

 pig were then worked very quickly, and a remarkably pure iron 

 was obtained, of which the following was the result of analysis : — 

 Combined carbon ... ... ... trace 



Silicon '005 



Manganese trace 



Phosphorus trace 



Sulphur '015 



This iron could only be forged in small pieces, even with the 

 greatest care, and therefore no results could be given as to its 

 mechanical properties. Dr. Hopkinson had determined the 

 magnetic properties of the metal, but the results are to be re- 

 served for the Royal Society. Speaking generally, it has 

 been found that the metal is more easily mai^netized for small 

 magnetizing forces than any other metal hitherto tested ; its 

 coercive force is less, its magnetization is greater, than any 

 other sample experimented with. The next .stage of Colonel 

 Dyer's experiments had for their object the utilization of ordi- 

 nary scrap steel, and the production, in the basic furnace, of steel 

 high in carbon and low in phosphorus, and at the same time to 

 decrease the wear and tear of the furnace. The principle of the 

 process consists in melting scrap with carbonaceous material, and 

 the results of the experiments have shown that when a pure 

 carbonaceous material and ferro-manganese free from phosphorus 

 can be obtained there will be no difficulty in producing a pure 

 carbide of iron containing only sufficient manganese for forging. 

 The author next described the method by which the process was 

 carried out. Nine consecutive charges were worked, with the 

 object of producing steel containing varying percentages of 

 carbon, to test the value of the process. The following table 

 gives the chemical analyses and the mechanical properties of the 

 steel of these charges : — 



Tbmsilk Tests. 



Chbmical Analysis. 



The paper by Messrs. Ball and Wingham, on the elimination 

 of sulphur, contained the results of experiments thoughtful and 

 suggestive in themselves, even if they do not show the iron and 

 steel maker any immediate results which he may apply. The 

 authors found that potassium cyanide placed on the surface of 

 molten cast iron almost completely removed the sulphur. Owing 

 to the extreme volatility of the cyanide, it was not found possible 

 to reduce the quantity required to within practical limits, and 

 efTorts were therefore made to find some flux which would retain, 

 when molten, a quantity of cyanide sufficient to effect the 

 desulphurisation. Sodium carbonate, lime, and blast-furnace 

 slag were in turn tried. It was found that the desulphurising 

 action was greater when the flux consisted mainly of sodium 



NO. I 179, VOL. 46J 



carbonate than when a less basic lime slag was used ; and that 

 in the latter case the diminution in the percentage of sulphur 

 varied directly with the amount of added cyanide. A table is 

 given of the results of the experiments, the best condition being 

 obtained when 200 grains of i-odium carbonate and ico grains of 

 potassium cyanide were used to 2000 grains of metal, when the 

 sulphur was reduced from 046 per cent, to o'o6 per cent. A 

 further experiment with sodium carbonate alone — 400 grains being 

 added to 4000 grains of metal — the sulphur was reduced from I'li 

 to 0*15 per cent. With caustic soda the sulphur was reduced 

 from o"i5 per cent, to 0*02 per cent., which is a satisfactory 

 result. The experiments also showed the facility with which 

 sulphur is reduced when present in large quantities, and that it is 

 the last part which gives difficulty in removing. Metallic sodium 

 was introduced into the bath in the form of an alloy with lead, 

 and this had the effect of entirely removing o'i8 per cent, of 

 sulphur. 



The paper by Mr. Reimers, which was the first read on 

 Friday, the second day of the meeting, does not call for notice, 

 excepting, perhaps, to remark that the Council of the Institute 

 were to blame for not taking care that the author was informed 

 beforehand that his contribution was not of a nature which 

 should have been submitted in the form in which it was read. Mr. 

 Callendar's paper on "Platinum Pyrometers" is a great con- 

 trast to the last-mentioned. The prominence given to the Le 

 Chatelier pyrometer in this country, by Prof Roberts-Austen 

 chiefly, has led to renewed hope on the part of those who desire 

 to measure higher temperatures. Mr. Callendar has been 

 amongst those who have been giving attention to the subject, 

 and the results of his labours, which are distinctly valuable, are 

 given in his paper, to which we would refer all practically 

 interested in the matter. His introductory remarks on air 

 pyrometers are interesting, and may be read with advant- 

 age by tho-e not already acquainted with this brancli 

 of the subject ; but it is of the Siemens electrical 

 resistance thermometer, known generally as the " platinum 

 pyrometer,'' that he has most to say. It has been hitherto 

 accepted that the platinum pyrometer was subject to the serious 

 defect of changing its zero with use. The British Association 

 Committee of 1874 discovered this, and it has since been amply 

 confirmed as a fact. The Committee experimented chiefly with 

 a pyrometer in an ordinary fire at moderate temperatures of 

 about 800° C, and they found that the resistance in- 

 creased continuously with heating, and that the wire underwent 

 rapid deterioration. They also made some experiments and 

 suggestions with a view to remedy this defect, but they did not 

 succeed in overcoming it. This continuous change of zero is 

 certainly the most serious practical defect that a pyrometer can 

 have, and there can be no doubt that the report of the British 

 Association did a great deal to destroy confidence in this method 

 of measuring temperature. 



We cannot do belter than continue Mr. Callendar's communi- 

 cation on this part of the subject in his own words : — 



" About seven years ago, when I began making experiments 

 on this subject at the Cavendish Laboratory, Cambridge, I was 

 at first very much surprised to find ttiat the platinum wires which 

 I used did not undergo continuous change, even when subjected 

 to much more severe tests than those applied to the Siemens 

 pyrometer by the Committee of the British Association. By 

 making further experiments, however, with a sort of imitation 

 Siemens pyrometer, I succeeded in reproducing at pleasure the 

 effects they had observed, and in proving to my own satisfaction 

 that these defects were not inherent to the method, but merely 

 incidental to the particular form of instrument on which they 

 experimented. I found that if the wire were properly protected 

 from strain and from contamination, the pyrometers could be 

 made practically free from change of zero, even at very high 

 temperatures. 



•• The construction of the Siemens pyrometer has not, so far 

 as I am aware, undergone any material change since 1874. The 

 coil of platinum wire, which forms the sensitive part of the 

 instrument, is wound on a clay cylinder, and packed in an iron 

 tube from 5 to 8 feet long, and about an inch or so in 

 diameter. I have here the fine wire and the clay cylinder from 

 a pyrometer which was recently in use at the Royal Arsenal, 

 Woolwich. I was informed that it had never been heated above 

 900° C, or iCoo° F., but its resistance had increased some 15 

 per cent., corresponding to an error of about ico° F. in the 

 temperature measurements. When the instrument was taken 

 to pieces it was found that the wire was quite rotten and brittle 



