KOnig.l ^14 [Jan. 19, 



but I have nowhere in the literature noticed a recognition of its superiority 

 over other methods, and especially none applying to this separation from 

 titanium and zirconium in a micro-chemical sense. 



The mother liquor from the alum, containing the sulphates of titanium 

 and zirconium, besides a considerable amount of ammonium sulphate, is 

 evaporated in a platinum capsule (one with a conical apex preferable, as it 

 facilitates the removal of the oxyds after ignition), evaporated to dryness 

 and ignited with caution over a blast lamp. The oxyds are then left in a 

 scaly condition, verj'^ finely divided and well adapted to the subsequent 

 treatment. They might be precipitated by ammonium hydrate and then 

 ignited ; but this would cause some inevitable loss, and on the other hand 

 silica would be again introduced by the filter ashes and from the glass 

 vessels, which would interfere with the reactions. The oxyds would be ob- 

 tained in hard pieces, requiring grinding, and augment the expenditure of 

 work and time generally. 



The difficulty, nay, impossibility, of separating titanium from zirconium 

 is well understood by all analysts who have occupied themselves with the 

 subject. I have not overcome these difficulties, but I hope being able to 

 show my co-laborers a way by which a quantitative estimation of the two 

 metals may be arrived at where thej' occur together. It was proposed 

 already by the great Nepom. Fuchs to reduce titanium dioxyd or chlorid 

 to the sesquichlorid by tin, and estimate its quantity volumetrically. Owing 

 to the unstable nature of the sesquichlorid and other causes not under- 

 stood, this method fails to give satisfaction. It is also proposed to effect 

 a separation by boiling the acetate solution of the metals. In my experi- 

 ments I have not been able to obtain anj'thing like constant results, and 

 the difficulty to remove the alkaline salts from the hj-dratic oxyds, causes 

 errors in small quantities of the oxyds, amounting to twenty-five per cent. 



In my prolonged working with titaninum and the earth of schorlomite, 

 I had occasion to produce the blow-pipe reaction many hundred times, 

 and noticing the various shades of color produced with microcosmic salt 

 in the reducing flame, together with their constancy, I conceived the idea 

 to make this reaction the basis of a quantitative colorimetric method. In 

 the first instance it should be useful in determinative mineralogy to dis- 

 tinguish between the several titanium minerals, where a difference of 10 

 per cent, even would be sutficient to characterize a species. In working 

 up the idea, however, I was astonished to see how rapidly the eye became 

 sensitive to discriminate between the shades and intensities of color on 

 small surfaces. Thus, when analyzing theastrophyllite, more minute and 

 careful experiments proved that a very satisfactory quantitative estimation 

 of titanium dioxyd maybe effected when it is mixed with zirconium or 

 other colorless oxyds, which themselves do not produce a coloration with 

 microcosmic salt in the reducing flame. There are but three metals 

 whose presence in even a comparatively small amount will interfere with 

 the accuracy of this method : vanadium, chromium and tungsten. The 

 two former produce a green salt in the reducing flame, the latter a deep 

 sky blue. 



