4 
pose. An order, to go to sleep again, elicited 
the question, “ If I go to sleep will it be dark 
again ? ” The child was inclined to regard the 
sleep and the darkness as a permanent sequence, 
and the tendency of the instinctive belief to 
come into play is manifest. Eut there is re- 
corded the yet stronger example, of a child 
asking why going to bed at night makes 
it grow light in the morning ? Here, 
the instinctive belief in the permanence 
of sequences had evidently misled the 
child. The inference is, that a thing in- 
stinctively believed to exist may yet have no 
existence, and that our instinctive beliefs, like 
our more complex beliefs, are liable to error. 
This instinctive belief, without which experience 
would be useless, this wondrous faculty, pro- 
phetic in its character, and enabling us to pre- 
dict the future from observation of the past, is 
a light to lighten our eyes, but not an infallible 
guide to our judgment. All that we can say of 
these beliefs is that they exist, and that the 
course of nature is in accordance with them. 
But this is a very different thing from saying 
that the laws of nature are immutable, and 
that their violation is physically impossible. 
Strictly speaking, the term impossibility should 
be confined to the domains of logic and mathe- 
matics. There is no such thing as physical 
impossibility. All physical knowledge is pri- 
marily (for I do not speak of deduced physical 
knowledge) derived from experience, and what 
a past experience has affirmed a future expe- 
rience may disaffirm. There is indeed one mode 
of seeming to establish the immutability, 
viz., assuming that future experience will 
always coincide in its results with past 
experience. But that is simply begging the 
question. The society will not misunderstand 
me, or imagine that I am calling in question the 
laws of nature, or the legitimate conclusions of 
physical science. My object is to guard against 
the laws of nature being supposed to teach 
what they do not teach, that is to say, their own 
immutability, necessity, and eternal duration. 
To say that they are immutable by any agency, 
is to say that science has disclosed to us all the 
agencies in the universe. To say that they are 
not merely existent and actual, but necessary, 
is to say that they are independent of experience 
and have the same characteristics as mathema- 
tical truth. But the reverse is the case. When 
we say that the three angles of a triangle 
are together equal to two right angles, or 
that a prime number can be found 
greater than any given number, however great, 
we say that which not only is true, but which 
cannot but be true, which is not derived from 
experience, which connot be demonstrated by 
any experience however numerous the experi- 
ments, which cannot be contradicted by experi- 
ence, and which is totally independent of ex- 
perience. Mark the contrast with the laws of 
nature, and consider one of the simplest of 
than. The child who regards sleep and dark- 
ness, or going to bed and daylight, as perma- 
nent sequences, does not err more than Aristotle 
in his gropings after that law now known 
as the first law of motion, and, like children, 
successive generations of men have to correct 
their first impressions by experience. All our 
knowledge of nature is derived from experience, 
and experience can only teach us what is, or 
has been, and what may and probably will be ; 
not what must be and cannot but be. Then, 
again, as to the eternal duration of the laws of 
nature. Here I shall quote from one who, to 
my mind, is the greatest metaphysician, as he 
certainly is one of the greatest mathematicians 
of this age. “ A clear view of the usages of 
nature must, of course, existing up to a certain 
point be augmented by reflexion, or further 
experiment, or both, up to a higher 
point; but no length of usage gives any 
odds in favor of the impossibility of the con- 
trary.” And, in a foot-note to another passage 
De Morgan says, “ If the laws of nature should 
continue unaltered until noon, the additional 
half hour will add a trifle to the force of their 
data. But the theory of probabilities, the only 
protector from false conclusions in such a 
case as the present, gives it as an un- 
doubted result that, no matter how many 
our observations of permanence from 
moment to moment may be, so long as they 
are finite in number, we cannot, from these 
observations alone, draw any probability, how- 
ever small, in favor of an unlimited continuance. 
Except by knowledge of continuance ah infinite ), 
we cannot acquire any well grounded faith in 
continuance ad infinitum , from any observation 
and reasoning grounded on t^iat observation 
alone.” I like the phraseology of De Morgan, 
and I think that the term usages of nature is 
at once more expressive and more accurate 
than laws, and less likely to conjure 
up the tiresome phantoms of immuta- 
bility, necessity and eternal duration. Let 
us be content to say, such and such are the 
usages of nature, and we believe them to be 
permanent for all the purposes of our experi- 
ence. But let us never lose sight of the fact 
that we have only moral, not mathematical, 
certainty of their permanence. Other than 
moral certainty is not, scientifically speaking, 
attainable unless with reference to a mathemati- 
cal or logical subject matter. And moral cer- 
tainty means no more than a high degree of 
probability, though it may sometimes be so 
nearly equivalent to mathematical certainty as 
to be arithmetically undistinguishable from it. 
I will give an illustration, but with 
greater circumlocution and complication than 
would be necessary if I were addressing an ex- 
clusively mathematical audience. And if to 
some it may sound strange to hear the laws of 
nature spoken of as highly probable sequences, 
they must remember that the probability spoken 
of is so high'.As to be properly termed moral, as 
distinguished from mathematical, certainty, 
