78 James Small. 
analogy of a piece of music with its main theme and the variations 
but applies it in rather a different way. He agrees with Bergson, 
however, in regarding evolution as a dual process. That the 
analogy of a stream cannot be followed without reserve is clear 
from the passage quoted at the beginning of Section B and from 
another which emphasises the part played by contingency in 
in evolution; “Two things only are necessary: (1) a gradual 
accumulation of energy; (2) an elastic canalization of this energy 
in variable and indeterminable directions, at the end of which are free 
acts.” 
If we place a number of water-falls and anastomosing back¬ 
waters at intervals along the metaphorical stream, we get some 
approach to the “explosions” and “free acts”. Bergson’s view 
of evolution can then be translated into scientific phraseology thus 
—the smooth flowing stretches of the “ stream ” correspond to 
orthogenetic development; the water-falls, explosions or free acts 
are saltations which give rise to the branches or back-waters of 
the stream; the riverbed with its sinuosities is the environment, 
which by epharmosis and elimination of the unfit keeps the stream 
within definite limits. In this way we get a synthetic, eclectic 
theory of evolution which includes all the diverse points of view 
and assigns to them their proper place in the complete scheme. 
In evolution by orthogenetic saltation *, with epharmosis and 
elimination of the unfit exercising a directing and delimiting 
function on the actual forms assumed by organised life 2 , we have 
the best of Darwinism, neo-Lamarckism, neo-vitalism, Mendelism 
and the mutation theory. 
Something of this kind, but without the precision rendered 
possible by recent work is foreshadowed by Bentham (cp. quotation, 
r Orthogenetic saltation may decide the inner constitution of a species, 
but the expression of that constitution is modified by epharmosis. Probably 
the inner constitution, as well as its expression in the form and structure of 
the plant, is altered by the cumulative effect of continued epharmonic adapta¬ 
tion The saltation may be either Mendelian or De Vriesian or both, Mendelian 
segregation having most evidence to support it as an originating cause of a large 
number of taxonomic species or even genera (cp. Rendle, 35). Progressive 
mutations seem to be as rare as the proved origin of new genes, but their 
rarity in quite in accordance with their place in the general scheme of 
evolution, as will be readily evident to anyone who tries to solve the 
Biochronic Equation of any highly organised species (see IV, 88, p. 674). It 
is remarkable how few characters which are not epharmonic are present in 
such a species and absent in others of lower systematic position. 
1 It is certain at least that epharmosis and, to a lesser degree, natural 
selection, decide the habit and many of the details of the structure of a species, 
even if the genetic constitution remains the same throughout a series of 
variations, a supposition which is, to say the least, unsupported by the 
necessary detailed evidence, 
