246 



SCIENCE. 



[N. S. Vol. V. No. 111. 



The best results seem to be obtained by 

 having the carbon in all steels, both stand- 

 ards and tests, in the condition given by 

 annealing, by having a number of standards 

 which differ little from each other in car- 

 bon content, and by not attempting to use 

 the method on steels containing very little 

 or very large amounts of carbon. It may 

 not be amiss to add here that the practice, 

 so prevalent in many of the steel works, of 

 using this method for all carbon determina- 

 tions, including those where contracts are 

 involved, is reprehensible and should be 

 discontinued. The chemist at the works 

 does the best he can with the method he is 

 using, and the amount of work required of 

 him, as well as the facilities furnished, do 

 not admit of the use of a better method. 

 On the other hand, when a dispute arises, 

 and it is ultimately shown that the works 

 are in error, the chemist is blamed and 

 analytical chemistry brought into disrepute, 

 not because either is really at fault, but be- 

 cause more is expected of the color test 

 method than it is really able to give. To 

 the steel makers we say, " Do not ex- 

 pect your chemist to render you the bricks 

 of good chemical analyses, unless you give 

 him the requisite straw of time and appli- 

 ances to do good chemical work." 



Few of the constituents of iron and steel 

 have more important influences on their 

 valuable qualties than phosphorus, and upon 

 few has more chemical work been done. 

 The present condition of the methods for de- 

 termining this constituent seems fairly sat- 

 isfactory, provided we are willing to take 

 time enough to do the work. In confirmation 

 of this statement, the work * done by the 

 Sub-committee on Methods of the Inter- 

 national Committee ou Standards for the 

 Analysis of Iron and Steel may be cited. 

 This sub-committee consisted of five mem- 

 bers, each of whom analyzed five samples of 

 steel, and each used his own method, with- 



*Proc. Am. Soe. dvUEng., 21, 59. 



out any attempt at consultation or agree- 

 ment with each other before the work was 

 done. The methods employed may be 

 briefly indicated as follows, those interested 

 being referred to the report of the committee 

 published as per the reference given for the 

 details. Mr. Blair used what is known as 

 the acetate method. Mr. Shimer used the 

 molybdate-magnesia method. Your speaker 

 used a combination of the acetate and 

 molybdate-magnesia method. Dr. Drown 

 used a combination of certain features of 

 the modern rapid methods of the molybdate- 

 magnesia method. And Mr. Barba on one 

 sample used the acetate method as de- 

 scribed by Blair, and on the other four 

 samples employed certain features of the 

 molybdate method to separate the phos- 

 phorous from the iron, and then used the 

 reductor to get the amount of phosphorus, 

 instead of weighing as magnesium pyro- 

 phosphate. It will be evident, to any one 

 carefully reading the report referred to, 

 that the methods employed differed widely 

 in principle, in strength of solutions and in 

 manipulation, and yet these methods gave 

 the following percentages of phosphorus in 

 the five samples : 



1. 2. S. 4. 0. 



Mr. W. P. Barba, 0.041 0.015 0.095 0.091 0.041 



Mr. A. A. Blair, 0.040 0.016 0.098 0.091 0.041 



Dr. T. M. Drown, 0.042 0.016 0.104 0.090 0.042 



Dr. C. B. Dudley, 0.040 0.016 0.099 0.097 0.039 



Mr. P.W. Shimer, 0.041 0.017 0.098 0.096 0.039 



In explanation of the results, we quote 

 from the report of the sub-committee: 



" Sample No. 1 is an ordinary open- 

 hearth steel. Sample No. 2 is a crucible 

 steel. Sample No. 3 is an open-hearth 

 steel to which metallic arsenic was added 

 while in the molten condition in a crucible. 

 Sample No. 4 is an ordinary Bessemer rail 

 steel. Sample No. 5 is the No. 5 sample 

 of the Committee on International Stand- 

 ards, and is an open-hearth steel. 



" It will be observed that the agreement 

 in the results on phosphorus obtained by 



