426 
CHEMISTRY: E. D. CAMPBELL 
observed. Solutions containing 70% of chlorhydrin can readily be 
obtained in this manner. 
It was also found possible to extract the chlorhydrin from its aqueous 
solutions by some immiscible solvent, benzene suggesting itself as the 
best from the practical standpoint. Thus, by a judicious combination 
of the three methods, — distillation, salting out and extraction, — chlor- 
hydrin could readily be obtained in any desired concentration and purity. 
1 The investigation was done under the auspices of the Bureau of Mines, War Gas Inves- 
tigations Department, December 1917-July, 1918. The paper in detail, approved for pub- 
lication by Major-General William L. Sibert, Director of the Chemical Warfare Service, U. 
S. A., will appear in the Journal of the American Chemical Society. 
2 Meyer, V., Berlin, Ber. D. chem. Ges., 19, 1886, (3260). Clarke, H. T., London, J. Chem. 
Soc, 101, 1912, (1583). 
STUDIES OF THE CONSTITUTION OF STEEL 
By Edward D. Campbell 
Chemical Laboratory, University of Michigan 
Communicated by M. Gomberg, July 30, 1919 
Researches on steel, considered as a solid solution, were carried on in 
this laboratory by the late J. Vv^. Longley prior to 1876. ^ Other investi- 
gations of iron and steel were carried on in the interval between 1876 
and 1891, at which latter time the present writer began a series of re- 
searches on the chemical constitution of the carbides of iron. The re- 
sults of these investigations have gone to show that the carbides found in 
steel have a much more complex structure than would be indicated by 
the commonly accepted formula FcsC. The idea that the carbides of 
iron and many other metals can best be studied if the carbides are re- 
garded as metallic substitution products of hydrocarbons was advanced^ 
in 1896. In addition to recognizing the complex molecular constitution 
of the carbides, the assumption has been made that the atomic relations 
existing between the carbides or other solutes dissolved in iron are es- 
sentially the same as those which exist between the molecules of sub- 
stances in aqueous solution and the water in which they are dissolved. 
Some further evidence in support of the hypothesis of the unity of 
mechanism of all solutions without regard as to whether the solvent is 
solid or liquid, metallic or a non-conductor of electricity, is given in two 
papers, one of which will be read at the Autumn meeting of the Iron and 
Steel Institute and the other at a forthcoming meeting of the Faraday 
Society. 
In the first of these it is shown that from twelve samples of steels, 
including straight carbon steel, high siHcon, high phosphorus, manga- 
