122 II. E. Merwin — Peroxidized Titanium Solutions. 



restores the color to a considerable degree to solutions bleached 

 by fluorine. Ferric sulphate produces effects similar to alumi- 

 nium sulphate, and also modifies the color because of the 

 color of its own solution. Phosphoric acid bleaches a standard 

 solution. Silica to the amount of *1 g. introduced in the form 

 of soluble sodium silicate into solutions considerably bleached 

 by fluorine, produced no more effect than could be accounted 

 for by the sodium sulphate generated. 



Doubtless there are many other substances that alter' the 

 color of peroxidized titanium solutions, but the ones here con- 

 sidered are the only ones ordinarily encountered in notable 

 amounts in solutions that would be used for the colorimetric 

 estimation of titanium or fluorine. 



Application to analytical processes. — In the estimation of 

 titanium by Weller's colorimetric method a correction must 

 be made for the effect of alkali salts, if such salts are present 

 in considerable quantity. The acidity of the solution must be 

 considered also. The above table shows the magnitude of the 

 corrections necessary for solutions containing *005 Ti0 2 , and 

 the amounts of free acid and of normal alkali sulphate indi- 

 cated. For example, 20 per cent too little, that is -004 g., of 

 Ti0 2 would be found in such a solution containing 6 g. of 

 alkali sulphate and 6 CC total free acid. 



For amounts of TiO y more or less than '005 g., these cor- 

 rections w r ill not hold. Twice this amount would require about 

 half the correction, and half this amount twice the correction. 

 Inasmuch as the correction can be made smaller by increasing 

 the acidity of the solution, it is highly desirable to do this. 

 In roclc analysis by using 6 g. of pyrosulphate, which is equiva- 

 lent to 4 g. of normal sulphate and 2 g. of acid, for the melt 

 containing the titanium, and dissolving this in water to which 

 10 cc of strong sulphuric acid has been added, a nearly negligi- 

 ble correction of only 3 per cent need be added. If the Ti0 2 

 exceeds *02 g. no correction is required. In case the melt is 

 dissolved in 100 cc of 5 per cent sulphuric acid, the titanium 

 found — if the amount is between -002 g. and '01 g. — is too 

 low by approximately -0004 g.* 



Estimation of Fluorine. — During the progress of this study 

 it was found that when the Nessler tube method was used, the 

 percentage ratios obtained by dividing the depth of each solution 

 bleached by fluorine by the depth of its matched standard, 

 could be plotted in lines so nearly straight that from the lines 

 simple formulas could be derived for use in analysis. f 



* It seems safe to conclude that the amounts of titanium in igneous rocks,, 

 estimated colorimetrically, have fallen short by nearly this amount in a 

 1-gram sample, for as much as 10 grams of pyrosulphate have often been 

 used, and probably seldom more than 200 cc of dilute acid. 



f The same ratios are obtained by dividing the final volume of the standard 

 by the volume of the test in cases in which a colorimeter is used which 

 requires the standard to be diluted. 



