February 12, 1897.] 



SCIENCE. 



249 



minations in a day. The accuracy of this 

 method has been questioned. "We are not 

 aware of any recent symposium on manga- 

 nese where different chemists using differ- 

 ent methods have worked on the same 

 steels. In our hands this method gives re- 

 sults closely agreeing with check work done 

 by the more laborious and generallj' ac- 

 cepted accurate methods, provided the sam- 

 ple contains not more than three-fourths of 

 a per cent. On samples containing over 

 one per cent, of manganese the results are 

 apt to be low, owing probably to the fact 

 that the manganese does not separate from 

 the nitric acid solution as manganese diox- 

 ide, but as some other oxide whose compo- 

 sition is not positively known. In the 

 calculation it is customary to regard the 

 separated oxide as manganese dioxide, and 

 this leads to perceptible error on large 

 amounts. Producers and consumers rarely 

 contend much over manganese in steel, and 

 methods for its determination have, perhaps, 

 not received, on that account, all the atten- 

 tion they deserve. There is evident need 

 of more work on this subject. 



The methods for the determination of 

 silicon can hardlj'' be regarded as in a per- 

 fectly satisfactory condition. If evapora- 

 tion to dryness, to render silica insoluble, is 

 employed, the time required is considerable. 

 If dehydration by means of sulphuric acid 

 and heat, as suggested by Drown,* is em- 

 ployed, there are difficulties which interfere 

 somewhat with accuracy. There seems 

 little doubt but that in skilled hands, with 

 sufficient care taken in the manipulation, 

 a couple of determinations may be made on 

 the same sample, using Drown's method, 

 that will agree closely with each other and 

 with results given by the longer and more 

 laborious methods. On the other hand,where 

 one operator is making a number of deter- 

 minations at the same time there is much 

 danger of error due either to failure to dehy- 



* Trans. Am. Inst. Min. Eng., 7, 346. 



drate sufficiently or to over-heating resulting 

 in the formation of insoluble iron salts. Our 

 experience indicates that the margin between 

 these two extremes is not very wide, and 

 that it is fully as frequent to have duplicates 

 on the same sample disagree as to agree. 

 Our observations point to the view that the 

 difficulty of insufficient dehydration is due 

 to the separation of iron salts as the sul- 

 phuric acid concentrates. These salts en- 

 close gelatinous silica and prevent the 

 dehydrating acid from getting at it. Unless 

 great pains are taken, therefore, to secure 

 this contact by sufficient stirring, the results 

 will be low. If, by some modification, the 

 iron salts could be kept in solution until 

 the silica is rendered quite insoluble it 

 would apparently be a decided step forward 

 with this method. It may not be amiss 

 here to call attention to the fact first 

 noticed in the laboratory of the Pennsylva- 

 nia Railroad Company,* that after the de- 

 hydration and subsequent dilution are 

 finished, if an interval of a few hours is 

 allowed to elapse before filtration, the silica 

 will redissolve and the results be low. Ap- 

 parently, as we are able to work the method, 

 the silica is not completely dehydrated, but 

 only sufficiently so that, if filtered at once, 

 fairly accurate results will be obtained. 



It is difficult to say anything positive 

 about the speed and output of Drown's 

 method. It is probably safe to say that a 

 couple of determinations could be made in 

 an hour and a half, but, on account of the 

 difficulty mentioned above, the method does 

 not lend itself well to working on a large 

 number of samples at once, and conse- 

 quently a large daily output is somewhat 

 interfered with. 



It must also be said of the methods for 

 the determination of sulphur in iron and 



* Address to the members of the Chemical Section 

 of the Engineers' Society, at Pittsburg, September 

 27, 1892, by C. B. Dudley, on ' Discrepancy in Chem- 

 ical Work by Different Workers. ' 



