Febeuaky 11, 1898.] 



SCIENCE. 



191 



use the pump to di'aw the liquid through 

 the reductor. This much beiug premised, 

 we may say that in a communication from 

 Mr. Porter W. Shimer, one of the mem- 

 bers of the Sub-Committee on Methods of 

 the Committee on International Standards 

 for the Analysis of Iron and Steel, he, 

 among other things, called attention to the 

 fact that when making a number of deter- 

 minations on the same sample, all other 

 things being the same, he got a reduced so- 

 lution that required more permanganate if 

 he prolonged the small tube below the stop- 

 cock in the reductor, nearly to the bottom 

 of the flask, than if this small tube pro- 

 jected only an inch or two into the flask. 

 This statement brought afresh to our minds 

 a thought that every one who has worked 

 much with molj^bdic acid must have had : 

 viz., that reduced motybdic acid is very 

 easily reoxidized. We accordingly deter- 

 mined to find out, if possible, whether this 

 was actually the case, and, if so, how much 

 this diiiiculty might amount to. Accord- 

 ingly, a stock solution of ammonia molyb- 

 date dissolved in water was prepared, and a 

 number of aliquot parts of this solution 

 measured out. Now, obviously, there are 

 two chances for the reduced solution to be- 

 come oxidized by exposure to the air. One 

 of these is from the air in the flask dur- 

 ing the reduction, and the other from the 

 outside air during the titration. Without 

 going into minute detail, it is, perhaps, 

 sufficient to say that when we reduced an 

 aliquot part of our stock solution, using the 

 short tube of the reductor and adding the 

 permanganate drop by drop, with continual 

 agitation during the whole titration, we used 

 22.7 cc. of our standard permanganate, all 

 figures given being a mean of a number of 

 closely agreeing determinations. When 

 now we made the reduction the same as be- 

 fore, viz. , with the short tube of the reduc- 

 tor, but titrated by allowing about ninety- 

 five per cent, of the permanganate required 



to run into the flask before agitation at all 

 and finishing the titration drop by drop, we 

 used 23.1 cc. of permanganate ; in other 

 words, so sensitive is a reduced solution of 

 molybdic acid that it is easy by varying the 

 mode of titration to introduce considerable 

 error. Prolonging now the tube at the bot- 

 tom of the reductor as suggested by Shimer, 

 which would result, as is apparent, in a 

 diminished exposure of the reduced solu- 

 tion to the air in the flask before titration, 

 we found our aliquot part to use up 

 23.6 cc. of permanganate. But even 

 with the prolonged tube there is some ex- 

 posure of the reduced solution to the air 

 during the reduction. Accordingly, on the 

 suggestion of my principal assistant, Mr. 

 F. ]Sr. Pease, we put a measured amount of 

 standard permanganate solution into the 

 flask which was to receive the reduced solu- 

 tion, more than sufficient to react with it, 

 and then prolonged the tube from the re- 

 ductor, to dip below the surface of this per- 

 manganate. Obviously, with this arrange- 

 ment the reduced solution is entirely 

 prevented from air exposure. On making 

 the reduction and titrating the excess of 

 permanganate with standard solution of 

 ferrous sulphate, it was found that the 

 aliquot part had now used up 24.1 cc. of 

 permanganate, an extreme difference in 

 amount of permanganate used under the 

 varying conditions described, of nearly six 

 per cent. Obviously, if two chemists were 

 working on the same sample of molybdic 

 acid, one employing the manipulation first 

 described, and the other that last described, 

 the discrepancy between them Would be 

 serious. The discrepancy on phosphorus in 

 steel, while the same in percentage, is very 

 much smaller in actual figures, but still 

 enough to be annoying. The work above 

 referred to is not yet quite finished, but 

 enough has already been done to demon- 

 strate that the ordinary method of deter- 

 mining phosphorus in steel can be advan- 



