338 Wisconsin Academy of Sciences t Arts, and Letters. 
The heat of formation of the chloride of the metal also does 
not- determine the action upon that metal, or even the order of 
violence of the action. 
The solubility of the chloride in the solvent used does not of 
itself enable prediction as to action upon the metal. This is 
shown by the experiment with benzene as solvent, in which zinc 
is attacked although its chloride is insoluble in benzene to any 
appreciable extent. Chemical individuality—the fact that dif¬ 
ferent substances act differently—appears vividly throughout 
this work. 
The tendency of the so-called new chemistry, the chemistry 
of infinitely dilute solutions and the gas equation PY = fRT, 
is to separate the particles of matter in solution so far from 
tions. The metals are acted upon, where action takes place, with great 
promptness, although slower action was noted in the case of some metals. 
The part played by protective coating on the metal in bringing the action 
to a stop is important. Even in solvents where the chloride of the metal 
is soluble, the solution becomes saturated with that metallic chloride, re¬ 
fuses to clear away the freshly formed chloride from the surface of the 
metal, and in consequence a coating is formed. The acid solution cannot 
get at the metal, and corrosion ceases. Mr. G. Ferneckes has recently 
shown that an aqueous solution of sodium hydroxide which has been sat¬ 
urated with the caustic will not act upon sodium continuously. A coating 
forms and the action ceases. (Part of an unpublished research now in 
progress in this laboratory.) 
The fact that speed of reaction is not proportional to the electrical con¬ 
ductivity has been noted in many cases. For example ammonium cyanate 
is transformed into urea 30 times faster in ethyl alcohol than in water, al¬ 
though the conductivity of the alcoholic solution is considerably lower 
than that of the aqueous solution. Certainly the degree of electrolytic 
dissociation here is not even a main factor in the reaction; yet on the 
basis of that theory we ought to find our main explanation in the activity 
of the ions. 
T. Erickson-Auren found that the speed with which zinc dissolves in 
aqueous solutions of HC1 and of SO s is not proportional to the electrical 
conductivity of these solutions; that presence of non-electrolytes in solu¬ 
tion lessens the speed of solution of zinc, but not in proportion to the mass 
of non-electrolyte added, rather in the proportion of the number of gram- 
molecules added; and that the presence of the zinc salt in solution in¬ 
creases the speed of reaction. (Zeit. Anorg. Chem. 27, 251, 1901.) He at¬ 
tributes this acceleration produced by the zinc salt to the increased con¬ 
ductivity of the solution; but when it appears that presence of zinc chloride 
in chloroform solution facilitates the corrosion of magnesium and alumi¬ 
num by dissolved HC1, it seems that this explanation is not sufficient. The 
conductivity of the chloroform solution containing ZnCl 2 and HC1 was 
approximately 30,000 times less than that of the 0.1. n HC1 in water with 
which Erickson-Auren worked. According to the idea of solution tension 
of metals, the more zinc chloride dissolved the less should be the solution 
tension, and the slower the solution of the metal. The facts indicate that 
this is not the case either in electrolytes or in non-electrolytes. The exist¬ 
ence of chemical affinity between the zinc and the solution seems to be the 
simplest explanation of the facts. If it be objected that the very object 
of our researches is to analyse chemical affinity, it may be said that a sat¬ 
isfactory explanation or analysis must fit the salient facts. 
