492 
On Scientific Method . 
[October, 
inferred the existence of more points of similarity ; they 
have then framed hypotheses which have guided them in 
their subsequent experimental investigations. To take an 
illustration : — When an eleCtric machine was worked a pe- 
culiar smell was noticed ; when a stick of moist phosphorus 
was allowed to remain exposed to air, a similar smell was 
perceived ; when a hot glass rod was dipped into a mixture 
of ether vapour and air, a similar phenomenon was percep- 
tible. From these observed similarities Schonbein inferred 
that the cause of the peculiar smell was probably the same 
in each case, and following up this analogy by experimental 
investigation he discovered ozone — a substance which has 
played, and is doubtless destined to play, a most important 
part in general chemical theory. 
Many instructive instances of the application of analogy 
are to be found in the science of chemistry ; in faCt that 
science is almost entirely founded on more or less general 
laws which have been deduced by analogical reasoning. 
The fadt that certain elements form groups having many 
common properties, and more or less sharply differentiated 
from other groups, has long been known. The further fadt 
that there is, in many instances, a regular gradation in the 
atomic weights of the members of such groups, seemed to 
point to a connection between atomic weight and general 
chemical behaviour of the elements. Many fadts were in 
keeping with this assumption. The connection between 
chemical properties and change in atomic weight has of late 
years been much attended to ; and it has been shown by 
Mendelejeff and others that, if the elements be arranged in 
order of their atomic weights, beginning with that which 
has the least atomic weight, the general properties — not 
only of the elements, but also of their compounds — may be 
regarded as functions of the atomic weights ; that, more- 
over, these functions are periodic, — that is, that groups of 
elements may be formed in order of increasing atomic 
weights, and that the general relations existing between, 
say, the third member of group two and the other members 
of the same group correspond with those relations which 
exist between the third member of group four and the re- 
maining members of this group. Following up the analogy, 
Mendelejeff has propounded an hypothesis which goes under 
the somewhat ambitious title of the periodic law, and from 
this hypothesis he has made certain predictions. Among 
other predictions he has foretold the existence of elementary 
bodies other than those we are acquainted with : he has 
even ventured to assign certain properties to some of these 
