1S93.] 81 [Hyatt. 
to me to explain why there are degenerative forms in the pliyhim 
which are indicated by the senile stages of the individual. 
The degenerative changes of the senile period may and |)racti- 
cally ill all cases do tend to the loss of the characteristics of the 
.adult period and consequently in extreme cases bring about not 
only the loss of a large proportion of progressive characteristics 
but loss in actual bulk of the body as compared with adults, as has 
been state<l above. This is usually regarded as due to the failure 
of the digestive organs or defective nutrition, and this may be true 
in many examples; but on the'other hand, it often begins in indi- 
viduals long before there is any perceptible diminution in size, and 
may occur in dwarfs and in some degenerate species in the earl}^ 
stages, and finally in series of species according to the law of tachy- 
genesis, so that one is led to believe that the tendency to the 
earlier inheritance of degenerative modifications producing retro- 
gression is inheritable like the tendency to the earlier inheritance 
of additional or novel chai-acteristics producing progression. Thus 
this law applied to progressive or x-etrogressive groups explains the 
mode in which their progression or retrogression is accomplished 
so far as the action of the laws of genesiology*are concerned. 
I have frequently tried to point out the obviously superfluous 
and erroneous character of the assumption that heredity should 
tend to act in two opposing ways at the same time. That it should, 
while bringing about or directly causing the perpetuation of like 
characters in endless chains of individuals, simultaneously produce 
or have a tendency to produce variation in any one of these links, 
is beyond belief. That it should act like gravity as a centripetal 
force holding the type true in the phylum and yet at the same time 
tend like a centrifugal force to make the same structures fly away 
from the type and put on novel characteristics, is obviously 
contradictory. 
There are, however, other causes for variations, among them 
the union of the two parents, which Weismann has too strongly 
accented. That new combinations of characters may be pro- 
duced in the young by the union of the characteristics, of the 
parents can hardly be denied even by those who are opposed to 
Weismann's mode of handling this class of phenomena. But 
that these combinations are important sources of origin for per- 
manent variations and that they make an impression upon the 
PROCEKDIXG8 i;: 3. X. H. VOL. XXVI. 6 AfGlST, 1893. 
