I'oulton.] 386 |Mayi6, 
dence required in order tli.vt there may be sufficient, depends upon 
the probability or improbability of the thing to be proved. This 
view is extremely well put by Professor Huxley in his memoir of 
TTume, where he says that if any one came to him and stated that 
he had seen a piebald horse in Piccadilly he would be prepared to 
believe it ; that he inight require confirmatory evidence if the 
statement were that a zebra had been seen ; but that if even the 
friend in whom he trusted told him he had seen a centaur trotting 
down that eminent thoroughfare, he should emphatically disbelieve 
it, and that nothing short of a monograph on the anatomy of the 
centaur by a comparative anatomist of the stamp of Johannes 
Miiller would convince him that the observation was correct. We 
are compelled to admit that the amount of evidence we require 
does to a great extent depend upon the inherent probability or 
improbability of the conclusion to be sustained. If it appears to 
us to be almost impossible to conceive of a mechanism whereby 
an acquired character can be transmitted from the outlying parts 
of the organism to its germ-cells, then we have every reason for 
scrutinizing most carefully any evidence that is alleged to prove 
such transmission. 
Let me first of all give you a concrete example which is fre- 
quently brought forward by those who believe in the Lamarckian 
theory in this country, and have chiefly studied the skeletons of 
Mammalia. They say the joint of an animal possesses just the 
sort of shape that would be produced by the motion of the joint 
itself, and they nrge that the joint as we see it has arisen from 
the hereditary effects of that motion. They look upon this as a 
very satisfactory explanation, because they consider it to be so 
obvious and fundamental. You do not require anything further, 
selection is unnecessary and even the individual variation — so 
mysterious a factor of the Darwinian theory — is here entirely 
explained. 
But is the interpretation valid? In the first place, it is clear 
that such an hypothesis can never afford a wide or general explana- 
tion. There are a great many parts of the animal body which 
are not modified in their use. You cannot thus explain the 
growth of hair, or the coloi- upon the surface of the organism. 
For these and other useful but ])assive structures, the Lamarckian 
interpretation will not hold at all. Hence we may divide the 
oro-anism into two sections, to one of which the Lamarckian 
