374 Ee de ALLEN: 
defects, however, persist in a latent condition in the germ plasm of these 
children, and if they are mated with those of like constitution the defec- 
tive characters all reappear in the grandchildren. The factors may even 
combine in such a way that some of these grandchildren are more defective 
than the defective ancestors from which they sprang. They unite the 
defects which were borne separately by the two grandparents. This no 
doubt explains some of the ill-effects which result from too close in- 
breeding. We should remember, too, that if here defects have been 
united, in other cases it would be equally possible that the excellences 
of different ancestors should be combined in some of their descendants. 
With Gammarus, however, what we have actually observed has been 
a degeneration of the eye, taking place step by step as one factor after 
another has been lost. Bateson, in his presidential address to the British 
Association in Australia in 1914, emphasised the fact that most, though 
perhaps not all, the Mendelian cases studied up to the present can be 
explained rather by the loss of factors than by the introduction of new 
factors. Since that address was delivered there has been, shall we say, 
in the air—for no one has ventured, I believe, to declare himself a com- 
plete adherent to ita theory of a kind of inverted evolution, starting 
with a highly complex primitive protoplasm or germ plasm, which by the 
loss of factor after factor has given rise to the endless varieties of plants 
and animals that we know. These factors are conceived of as being for 
the most part restraining or inhibiting factors, whose loss, one by one, 
the course of ages has allowed the full powers and glories hidden in the 
primitive plasm to unfold themselves—a process which still goes on. 
What the final excellence or final catastrophe is to be, when all the bonds 
are broken and all the restraints are lost, no one, as far as I know, has 
ventured to suggest. When, however, we take into consideration the 
whole range of facts upon which our conceptions of organic evolution are 
based we find little to support such a view. 
The cases of Mendelian inheritance which I have so far discussed have 
been of a simple character, following exactly the law which Mendel first 
laid down. Sometimes, however, the phenomena are more complicated. 
We saw that the albino-eye of Gammarus was always imperfect in shape. 
Absence of colour and imperfect form are here always united and remain 
united in inheritance. Characters which behave in this way are spoken 
of as linked characters, and the factors in the germ cells from which they 
originate are also said to be linked. There is often also a special connec- 
tion between a particular character and the sex of the animals which 
transmit and inherit it. This is known as sex-linkage and is well illus- 
