334 
for the benefit of the whole structure. These phenomena 
are characteristic of all specific types, and consequently 
the transmutation of species involves the change of a 
structure of co-ordinated parts into another type of organ 
ism whose parts are co-ordinated in some other way. 
But this is not all. It would seem that there is a tend 
ency inherent in these organisms which assists in this 
transmutation. It is expressed by saying that “when one 
part is modified . . . other parts of the organisation 
will be unavoidably modified.”* This principle is best 
described as the law of correlated variation. The various 
manifestations of this principle are due to three causes— 
symmetry, function, and development. It is clear that if 
one symmetrical organism is to be changed into another, 
the change of one part will necessitate the change of 
others also. It is equally evident that if the new organ 
ism is to be an efficient apparatus, the change in one 
part must be accompanied by others also. This a priori 
necessity is realised in the actual world. Structures which 
have obviously undergone modification are still charac 
terised by symmetry of form and the efficient co-operation 
of parts. 
As an illustration of the influence of symmetry, we may 
take the case of the ancon ram, in which we have a birth 
variation amounting to a new variety, if not to a new 
species. It is reasonable to suppose that in such a case 
pressure on one part may have modified other parts, so 
as to substitute one proportion of parts for another; for, 
as Mr. Darwin says :— 
“A modification which arises during an early stage of growth 
tends to influence the subsequent development of the same part, 
as well as of other and intimately connected parts.”—(The Variation, 
vol. ii., p. 320.) 
* Darwin. The Variation, vol. ii., p. 320.