Validity, and Residual Affinity. 303 



bably in many cases molar, physico-chemical experiments are 

 forcing npon bewildered chemists. We must at any rate 

 leave many of them in quarantine for some time to come. 

 And, considering that the mortal chemist will always have to 

 do almost all his work at the bottom of a varyingly moist 

 aerial ocean, it is surely time that a protest was made against 

 the claims for admission to the chemical record of so-called 

 " new " bodies, which are simply more or less moist than the 

 ordinary substance. Messrs. Carnelley and Walker have 

 recently carefully * adduced instances of these chemical im- 

 postures. It is quite refreshing to find among them an honest 

 hydrate like Ce(OH) 4 which can stand 600° f ; but surely the 

 line should be drawn at the temperature of hot water. In 

 other cases of alleged " molecular compounds " some tangible 

 chemical evidence should be required of their existence, 

 or at least of the chemical homogeneity of the material 

 or watered substance which is said to contain them. At the 

 most, then, the so-called " silver hydrate " should be formulated 



Ag — — Ag 

 as and not as AgOH. To use the latter and to 



HOH 

 use HC1 for ordinary " hydrochloric acid " is to strain out the 

 gnat and swallow the camel. The fact probably is that it has 

 become popular in preference to Ag 2 + OH 2 , because of the 

 frequent replacement of CI by OH through its use in organic 

 chemistry. Even in using the formula HC1 conjointly with 

 NH 4 (0H), we are guilty of going against the weight of evi- 

 dence, probably influenced, as Prof. Armstrong (rightly as it 

 seems to me |) suspects, by the " ammonium ,} theory. Liquid 

 hydrochloric ' ' acid " does not dissolve a single metal, except 

 perhaps aluminium. And it may be asked, Why should a 

 liquid metallic chloride be expected to do so ? The common 

 equation Zn + 2HCl=ZnCl 2 + H 2 is, of course, quite untrue. 

 It is even more objectionable than the misleading sheet- 

 anchor of so many textbooks : Zn + S0 4 H 2 = ZnS0 4 + H 2 . In 

 " hydrochloric acid " the chlorine is probably tervalid and the 



* Journ. Cheni. Soc. Jan. 1888, p. 59. 



f This is probably connected with the quite exceptional volatility of 

 osmium oxide. The metallic atom is possibly inside an oxygen nucleus, 



HO- — -OH 



as it were, | /Os(| . 



HO^ ^OH 



X This is instanced in the ingenious [NH 2 (Hg 2 )"]' formulae given in 

 the case of mercur-ammonium and other difficult compoimds. Of course 



Cl=Hg Cl=Hg" 



if I be admitted for calomel, | follows as a corollary. 



Cl=Hg NH 2 Hg" 



