254 
THEORIES OF HEREDITY. 
Nov., 1889. 
individuals are derived from “identical” ova, the above- 
mentioned arguments and conclusions will receive very strong 
support. 
“Identical” twins are invariably of the same sex. When 
twins are of different sex, the degree of resemblance is no 
greater than that between brothers and sisters generally. 
This is also true of many twins of the same sex, and Gallon 
lias brought forward evidence to show that they may even 
differ more widely than is usual with brothers or sisters. 
It has been long known that twins of the same sex are 
often enclosed in the same embryonic membranes, while; 
twins of opposite sex are always enclosed in separate mem¬ 
branes. The latter would be the product of distinct ova, 
which had been separately fertilised, as in the ordinary 
multiple births of animals (cats, dogs, rabbits, Ac.). The 
former would be the product of a single ovum, which has 
divided into two ova, in all probability after fertilisation. 
But it is clear that the ova arising from the two halves of a 
single ovum, at a time when the individual characteristics 
were already determined, would be very nearly identical: 
their resemblance would be of a very different order to that 
of separate ova. We also find that twins of the same sex 
present resemblances of a very different order to that of 
brothers or sisters generally, who are developed from separate 
ova. It must be admitted, therefore, that there is a very high 
degree of probability that the “ identical ” ova are those which 
develop into the “ identical ” individuals. The interesting 
conclusion that sex is also predetermined in the fertilised 
ovum also follows from the same facts. 
The probable beginning of the development of “identical ” 
twins is shown in Diagram III. A* is a fertilised ovum with 
the individual characteristics predetermined ; it is, therefore, 
different from A in Diagram II., and is distinguished by the 
asterisk. At its first division A* does not form the cells of 
Stage II., which, it will be remembered, are different from 
each other and from the ovum ; but it divides without differ¬ 
entiation into two equivalent cells, like each other and like 
the ovum. Hence the first division of A* does not produce 
Stage II., but Stage Ia, consisting of two similar ova. 
Each of these then divides, as in Diagram III., forming a true 
Stage II., comparable to that of Diagram II. After this the 
other stages succeed as in the latter, and finally two individuals 
will be formed, which must resemble each other if it be true 
that individual characteristics are predetermined in the fer¬ 
tilised ovum. And, as a matter of fact, such resemblances are 
seen in individuals whose development may be considered, 
