1877.] ^1^ [Konig. 



will be published hereafter iu connection with the reactions of the prob- 

 lemalic earth in schorlomite.) If the solution be very neutral, a large per- 

 centage of aluminum is precipitated. Owing to this behaviour, titanic 

 dioxyd is often given as alumina, and vice versa. When the ferric oxyd, 

 after the supposed precipitation of titanium, is weighed, re -dissolved and 

 estimated by permanganate, the deficit is considered to be alumina, when 

 in reality it is titanium, or zirconium dioxyd. However, my method 

 although applicable to large quantities is more especially designed for small 

 residues, with which precipitations, for physical and chemical reasons, give 

 rise to considerable error. 



The mixed oxyds are digested with equal parts of water and concen- 

 trated sulphuric hydrate at increasing temperature, until the excess of acid 

 is evaporated. I dissolve now in the smallest possible quantity of water, 

 and bring the liquid upon a watch glass, where a quantity of diammonium 

 sulphate is added, corresponding to the alum equivalent of alumina equal 

 to the weight of the mixed oxyds. A gentle heat is applied until the am- 

 monium salt is dissolved, and the whole is allowed to evaporate spontan- 

 eously in a warm place. Under these conditions, titanium, zirconium and 

 aluminum form double salts with ammonium, but those of the two former 

 are very soluble in cold water ; while the well-known aluminum double 

 salt, the alum, partly owing to the physical conditions of the crystals, is 

 only slightly soluble in cold water, and still less so in a solution of ammo- 

 nium or potassium sulphate. Accordingly, the dry mass of salts is moist- 

 ened with a few drops of cold water, when every salt, but the alum will 

 pass into solution, and the latter salt will be recognized at once, under the 

 lens, by peculiar and constant octahedric forms. 



This salt is so granular, that it can readily be freed from accompanying 

 salts by decantation, washing with a solution of ammonium sulphate. Com- 

 parative tests showed tliat this washing process may be continued at length, 

 without effecting a perceptible decrease of the alum. Thus freed from its 

 mother liquor, the crystals may be removed into a small platinum crucible 

 by means of a small piece of filter paper, ignited over the blast lamp, 

 aluminum sesquioxyd obtained in great purity, and its weight ascertained. 

 If the quantity be very small (less than 0.0005 gr. ) the filter paper, whose 

 ashes bring in impurity, the crystals may be washed into the crucible with 

 a small amount of boiling water. In this manner I could estimate 0.0001 gr. 

 and obtain a very distinct confirmative test by cobalt nitrate, before the 

 blowpipe. Experiments demonstrate that the presence of titanium and 

 zirconium sulphates does not interfere with the complete crystallization of 

 aluminum as alum, and as this salt only contains eleven percentum of 

 aluminum sesquioxyd, the sensitiveness of the reaction is readily under- 

 stood. 



Note. — The mixed oxyds must be freed from silica, which is easily ac- 

 complished by adding a drop or two of hydrofluoric acid, during their diges- 

 tion with sulphuric acid. 



I am well aware that the determination of aluminum as alum is not new, 



