Nov., 1889. 
THEORIES OE HEREDITY. 
253 
process. Many facts, however, seem to show that the 
principle of development is correctly indicated in the diagram. 
At about the time of the last division we must suppose 
that the minute mass of germ-plasm, A, grows and separates 
as a germ-cell or germ-cells from either FI or one or more of the 
somatic cells into which the latter divides. The four germ- 
cells of the adult Metazoon are then produced by division. These 
germ-cells are. therefore, similar to that which started develop¬ 
ment ; they are, in fact, a piece of it, which has grown without 
undergoing anv essential alteration. The develonment of 
these four germ-cells will, therefore, produce offspring 
resembling their parents. 
If, however, some of the somatic cells become modified 
from that nature which was predetermined in the germ-plasm 
of the ovum, there is no way in which the hereditary trans¬ 
mission of such effects can be explained by the theory of the 
continuity of the germ-plasm ; for the theory does not include 
anv means by which the effects could be conveyed to the 
germ-cells, or, if conveyed, could produce in them changes 
such that similar effects would be predetermined in the corre¬ 
sponding somatic cells of the offspring. The acquired 
changes in Ir and Nl, indicated by their dark colour, would 
be confined to the organism in which they arose, and would 
not affect its offspring, at any rate in a corresponding manner. 
If the transmission of acquired characters w r ere proved to 
be an undoubted fact, Weismann’s tlieorv of hereditv would 
inevitably collapse. It cannot, liow r ever, be maintained that 
such proof is yet forthcoming. 
The question largely turns upon an exact knowledge of 
the proportion borne by blastogenic to somatogenic characters. 
We know how important a share of our physical and mental 
qualities are hereditary. It would, therefore, follow, if Weis¬ 
mann’s theory be correct, that blastogenic characters are far 
more important than somatogenic. 
There is some evidence that this is the case, and I will 
here bring forward one line of proof, because it also supports 
the conclusion alluded to above, that the whole organism is 
predetermined in the ovum. 
If this last conclusion be valid, it follows that the differ¬ 
ences wdiicli characterise individuals are predetermined in the 
ova from which they arise, that ova are not alike any more 
than individuals. We do, however, occasionally meet with 
individuals so much alike that we (incorrectly) speak of them 
as “identical.” The resemblance between certain twins is 
far closer than that between other members of the same 
family. If, therefore, we can prove that such Si identical” 
