complex molecule. The evidence thus gained points to
the conclusion that, out of some primordial units, the so-called
elements arise by compounding and recompounding; just as by the
compounding and recompounding of so-called elements there arise oxides,
and acids, and salts.

And this hypothesis is entirely in harmony with the phenomena of
allotropy. Various substances, conventionally distinguished as simple,
have several forms under which they present quite different properties.
The semi-transparent, colourless, extremely active substance called
phosphorus may be so changed as to become opaque, dark red, and inert.
Like changes are known to occur in some gaseous, non-metallic elements,
as oxygen; and also in metallic elements, as antimony. These total
changes of properties, brought about without any changes to be called
chemical, are interpretable only as due to molecular rearrangements;
and, by showing that difference of property is producible by difference
of arrangement, they support the inference otherwise to be drawn, that
the properties of different elements result from differences of
arrangement arising by the compounding and recompounding of ultimate
homogeneous units.

Thus Mr. Martineau's objection, which at best would imply a turning of
our ignorance of the nature of elements into positive knowledge that
they are simple, is, in fact, to be met by two sets of evidences, which
imply that they are compound.

* * * * *

Mr. Martineau next alleges that a fatal difficulty is put in the way of
the General Doctrine of Evolution by the existence of a chasm between
the living and the not-living. He says:--"But with all your enlargement
of data, turn them as you will, at the end of every passage which they
explore, the _door of life_ is closed against them still." Here again
our ignorance is employed to play the part of knowledge. The fact that
we do not know distinctly how an alleged transition has taken place, is
transformed into the assumption that no transition has taken place. We
have, in a more general shape, the argument which until lately was
thought conclusive--the argument that because the genesis of each
species of creature had not been explained, therefore each species must
have been separately created.

Merely noting this, however, I go on to remark that scientific discovery
is day by day narrowing the chasm, or, to vary Mr. Martineau's metaphor,
"opening the door." No