620 REPORT—1899, 
fact that in very thin layers it ceases to behave as in thicker layers. But when 
we pass on from this general statement and give definite form to the granules or 
assume definite qualities to the intergranular cement we are dealing with pure 
hypotheses. 
It is hardly possible to think that we shall ever see an atom or handle the 
ether. We make no attempt whatever to render them evident to the senses. We 
connect observed conditions and changes in gross visible matter by invisible mole- 
cular and ethereal machinery. The changes at each end of the machinery of 
which we seek to give an account are in gross matter, and this gross matter is our 
only instrument of detection, and we never receive direct sense impressions of the 
imagined atoms or the intervening ether. ‘To a strictly descriptive physicist 
their only use and interest would lie in their service in prediction of the changes 
which are to take place in gross matter. 
It appears quite possible that various types of machinery might be devised to 
produce the known effects. The type we have adopted is undergoing constant 
minor changes, as new discoveries suggest new arrangements of the parts. Is it 
utterly beyond possibility that the type itself should change ? 
The special molecular and ethereal machinery which we have designed, and 
which we now generally use, has been designed because our most highly developed 
sense is our sense of sight. Were we otherwise, had we a sense more delicate 
than sight, one affording us material for more definite mental presentation, we 
might quite possibly have constructed very different hypotheses. Though, as we 
are, we cannot conceive any higher type than that founded on the sense of sight, 
we can imagine a lower type, and by way of illustration of the point let us take 
the sense of which my predecessor spoke last year—the sense of smell. In us it 
is very undeveloped. But let us imagine a being in whom it is highly cultivated, 
say, a very intellectual and very hypothetical dog. Let us suppose that he tries 
to frame an hypothesis as to light. Having found that his sense of smell is excited 
by surface exhalations, will he not naturally make and be content with a corpus- 
cular theory of light? When he has discovered the facts of dispersion, will 
he not think of the different colours as different kinds of smell—insensible, 
perhaps, to him, but sensible to a still more highly gifted, still more hypothetical 
dog ? 
Of course, with our superior intellect and sensibility, we can see where his 
hypothesis would break down; but unless we are to assume that we have reached 
finality in sense development, the illustration, grotesque as it may be, will serve . 
to show that our hypotheses are in terms of ourselves rather than in terms of 
Nature itself, they are ejective rather than objective, and so they are to be regarded 
as instruments, tools, apparatus only to-aid us in the search for truth. 
To use an old analogy—and here we-can hardly go except upon analogy— 
while the building of Nature is growing spontaneously from within, the model of 
it, which we seek to construct in our descriptive. science, can only be constructed 
by means of scaffolding from without, a scaffolding of hypotheses. While in the 
real building all is continuous, in our model there are detached parts which must 
be connected with the rest by temporary ladders and passages, or which must be 
supported tili we can see how to fill in the understructure. To give the hypo- 
theses equal validity with facts is to confuse the temporary scaffolding with the 
building itself. 
But even if we take this view of the temporary nature of our molecular and 
ethereal imaginings, it does not lessen their value, their necessity to us. 
It is merely a true description of ourselves to say that we must believe in the 
continuity of physical processes, and that we must attempt to form mental pictures 
of those processes the details of which elude our observation. For such pictures 
we must frame hypotheses, and we have to use the best material at command in 
framing them. At present there is only one fundamental hypothesis—the mole- 
cular and ethereal hypothesis—in some such. form as is generally accepted. 
Even if we take the position that the form of the hypothesis may change as 
our knowledge extends, that we may be able to devise new machinery—nay, even 
that we may be able to design some quite new type to bring about the same 
