722 PRELIMINARY OBSERVATIONS. 



of the original compound. The equivalent substitution of 

 chlorine for hydrogen is thus admitted in its full extent. This 

 substitution appears to confirm and extend what is generally 

 understood at the electro-chemical distribution of elements, on 

 which our ideas of binary combination are founded, and not 

 to be opposed to or incompatible with such views. 



The isomorphism of chlorine and hydrogen, equivalent for 

 equivalent, if confirmed, will cause the removal of the latter 

 element from the magnesian class, but establishes a relation 

 between hydrogen and silver, and therefore places the former 

 on the verge of the potassium group. Hydrogen certainly possesses 

 many relations to mercury, if not absolutely isomorphous with 

 that metal (page 656). 



By a chemical type M. Dumas understands a certain number 

 of elements combined together, every one of which, whatever 

 be its nature, may be replaced by another, and indeed every 

 one in its turn, so that not a trace of the original compound 

 may remain. The arrangement of the elements in regard to 

 each other remains always the same, and that is the character of 

 the type. Thus all the chlorinated compounds derived from 

 hydrochloric ether, including perchloride of carbon, contain ten 

 atoms, which are supposed to have the same arrangement as 

 in hydrochloric ether itself. Aldehyde produced by the partial 

 oxidation of alcohol, and chloral by the action of chlorine upon 

 alcohol, contain the same number of atoms and belong to one 

 type: 



Aldehyde, C 4 H 3 O + HO, or C 4 H 4 O 2 . 

 Chloral, C 4 C1 3 O + HO, or C/j^CV 



When on the other hand, an atom is withdrawn from a 

 compound without being replaced by another, the atoms which 

 remain cannot retain their original position, and a new type 

 must result. It is supposed by Dumas, that chlorine may replace 

 carbon, nitrogen and other elements besides hydrogen, and any 

 one element, indeed any other, without destruction of the 

 primitive type. Such substitutions however, have not been 

 effected, and on the view of substitution taken above are not 

 to be expected, for carbon at least must always be considered 

 as the basyle or electro-positive constituent of organic com- 



