283 
our experience, and proper inferences from it ; and for any- 
thing we could tell a 'priori they might all have been different. 
That great saying of Sir J. HerscheTs should never be for- 
gotten, that a sufficiently clever man shut up by himself 
might conceivably reason out all mathematical truth, up to the 
highest that will ever be reached ; but the cleverest man that 
ever lived could not divine a priori how a lump of sugar 
would behave when put into a cup of tea. There must also 
be laws of nature of which we yet know nothing more than 
that they are wanted to explain some phenomena of which we 
know no cause. A constant phenomenon can only be regarded 
as itself a law of nature, until some cause behind it is dis- 
covered, which then takes its place. Some physiological 
phenomena are variable and uncertain, such as the different 
effects of the same food and medicines on different persons, 
though they are all doubtless in conformity with some law. 
The still more precarious phenomena of mesmerism can 
neither be ignored or got rid of by any rational hypothesis, 
however often they are tainted with fraud ; or of occasional 
apparitions, and perhaps a few kinds of divination, which are 
all beyond the reach of any law that is yet known or imagined. 
All that is quite apart from Miracles, of which I have nothing 
to say here, especially as I have treated of them in a lately- 
published S.P.C.K. tract, called “A Review of Hume and 
Huxley on Miracles. ” 
The argument of the “ Origin of Laws of Nature” is, that 
the only alternatives for cosmogony are, (1) a single Creator 
who made and maintains the laws of nature; and (2) as 
many creators as the atoms. of the universe, all agreeing how 
they would behave, and always keeping their resolutions ; 
and they must also have had foresight enough to agree on the 
laws of nature, or of their respective motions, that would 
produce all the actual results. As that alternative is hardly 
possible for any rational man to accept,* it necessarily 
follows that between those two the other is the true one, 
viz., that there was one Creator; and a Creator omnipotent 
enough to make all the laws of nature must, a fortiori , 
* And yet I see, from Mr. Goldwin Smith’s article on Mr. Leslie Stephen 
and Herbert Spencer in the Contemporary Review of last December, that 
some philosopher, whom he does not name, has accepted this “ pan-atomic ” 
theory as the only logical alternative to a Creator. So far that philosopher 
is quite right, and it is satisfactory to see it acknowledged. [Nevertheless, a 
newspaper critic of this lecture said it was absurd to state such an alterna- 
tive : so much he knows about it.] 
Laws of 
N ature are 
not Axioms. 
