440 RECORD OF SCIENCE FOR 1887 AND 1888. 



somewhat from those which have? becu hitherto generally held, Aftei' 

 arguing" iu favor of simplicity of strncture, he discusses a number of the 

 common aluminous silicates, and shows that the, more complex of them 

 are easily represented as substitution derivatives of normal salts, alu- 

 minum itself being regarded as trivalent. Thus, from the normal ortho- 

 silicate of aluuiinum, xenolite, the species tibrolite, topaz, muscovite, 

 paragonite, eucryptite, dumortierite, grossularite, prehnite. and natro-- 

 lite are simply derived, and in away which accords well with their 

 actual occurrences and alterations in nature. The paper concludes 

 with the general hypothesis that all double salts are substitution deriv- 

 atives of normal compouuds. (Am. Chem. Journ., x, 120.) 



In a later number of the same journal (x, 405) E. A. Schneider shows 

 that the differences in the behavior of silicates toward hydrtchloric acid 

 are in the same order as the differences pointed out in Clarke's formula. 



The constitution of the clays. — By a careful study of the thermal phe- 

 nomena attending the dehydration of many clays, Le Chatelier arrives 

 at the conclusion that they represent only Ave compounds, as follows: 



Halloysite 2 SiO^AljOs . 2H,0 . aq. 



Kaolin 2 SiO.2 AljOs . 2H2O 



AUopbaue SiO-j AljOi . aq. 



Pyrophyllite. 4 SiO. Al.Os . H2O 



Moutmoiillonite 4 SiO^ AI2O3 . H-O . aq. 



Of these the last one is doubtful. (Zeitschr, Phys. Chem., i, 396.) 



Hydroxychlorides of titanium. — According to Koenig and Otto von der 

 Pfordten, dry, gaseous hydrochloric acid does not act upon titanium 

 tetrachloride. On the other hand, concentrated aqueous hydrochloric 

 acid acts upon the chloride, forming a series of hydroxy-derivatives, 

 which may be represented as follows: TiCl4; TiClsOH; TiC]2(CHj2; 

 TiCl(0H)3. These compounds are called by the authors i'.hlorides of 

 titauic acid, and all are solids. The monochloride is stable in the air 5 

 the dichloride is very hygroscopic; the trichloride somewhat so. (Be- 

 richte, xxi, 1708.) 



Titanic acid in soils. — J. F. McCaleb has examined fourteen samples 

 of soil froui central Virginia, together with one from Alabama and one 

 from Nevada, findiug titanic oxide present in all in percentages ranging 

 from 0.33 to 5.42. Although this wide distribution of titanium was rea- 

 sonably to be expected, it seems not to have been hitherto observed. 

 (Am. Chem. Journ., x, 36.) 



Germanium. — Several jjapers upon this interesting new metal have 

 appeared, which identify it completely with the predicted "ekasilicium" 

 of Mendelejeff. For the specitic heat of the metal, between 0° and 

 440°, Nilson and Pettersson hud 0.0757, which gives an atomic heat of 

 5.33. They also find normal values for the vapor density of GeCl4, Gel4, 

 and GeS. (Berichte, xx, ref. 134.) 



By Winkler several compounds of germauium are described, notably 

 a germanium chloroform, GeHCla; germanium ethyl, Ge(02tJ5)4j and 



