THE OLD AND THE NEW CHEM1STKY. 
31 
The entire series of elements, arranged as described, may be 
subdivided into sections, and these sections are generally com- 
posed of elements bearing a close chemical analogy to one another. 
The valency of the numbers of many subsections increases and 
decreases periodically with changes in combining number 
and in specific volume. But it would be beyond the scope 
of this paper to discuss the bearings of the “ periodic law of 
Mendelejeff” in detail. Suffice it to say that the “periodic 
law ” evidently points to a quantitative connection between the 
combining numbers of the elements and the chemical (and 
many of the physical) properties of these bodies, and of their 
compounds.* 
In the foregoing statements the combining numbers of the 
elements must be understood to mean those numbers which 
are based upon the two-volume formulae of series of compounds, 
and which are checked by specific heat determinations, and by 
considerations of crystalline form. 
The old equivalent numbers of the elements may be arranged 
in ascending order, but no general periodic connection can be 
traced between these and the properties of the elements. 
The old chemistry refuses such generalisations as those which 
I have briefly indicated. It will have no dealings with two- 
volume formulae, with numbers deduced from specific heat de- 
terminations, or from isomorphism ; it traces no periodic connec- 
tion between the numbers which it assigns to the elements, and 
the general properties of those elements. These things are 
called, by the old chemists, hypotheses ; their system is founded 
on facts, and on facts alone. I suppose they mean to imply a 
very grave censure by the use of that word hypothesis. They 
seem to forget that all great advances in science have been 
made by the proper use of hypotheses. 
But the old system does not confine itself to facts ; else, why 
does it adopt 14 as the equivalent of nitrogen? why 6 as 
the equivalent of carbon? Why does it unite the equivalent of 
aluminium sulphate as A1 2 0 3 .3S0 3 , when the true equivalent 
amount of this body is represented by the formula Al s 0.S0 3 ? 
The old system deals in hypotheses as well as the new. There 
is, it is true, a difference between the hypotheses : the hy- 
potheses of the old school are indefinite ; they escape one when 
one attempts to lay hold of them ; they do not admit of definite 
deductions being made from them. The hypotheses of the new 
chemistry are, for the most part, definite, workable, and allow 
of deductions being made from them, which deductions may 
be tested by an appeal to facts. 
* In a paper entitled 11 On Chemical Classification ” in the Philosophical 
Magazine for August, September, and October, 1877, I have gathered to- 
gether in some detail the facts concerning this connection. 
