APPENDIX. 
131 
modifications were in some way directly transmitted to offspring. 
But Weismann says “ No ; modifications of structure caused by 
environment cannot be inherited, because the germ plasm is 
not derived from the body, but the body from the germ plasm.” 
What, then, is the cause of variation in the germ plasm ? Weis* 
mann says that the complicated constitution of the germ plasm is 
the cause of a condition in it of unstable equilibrium of nutrition, 
and, consequently, of slight irregularity of growth, and that this 
is the cause of continual minute variations in the -structure 
of the germ plasm. These minute variations do not as a rule 
cause any perceptible change in the structure of the body, but 
when a number of them happen to become superposed one upon 
the other we get an accumulation, as it were, of slight modifica- 
tions resulting in a perceptible variation. It is as though we had 
a sheet of water with very slight ripples upon the surface. So 
long as these ripples follow each other in regular succession they 
are scarcely noticeable, but if it should happen that three or four 
ripples coincide in point of time and place, then we should get a 
noticeable wave. 
Having now given you a very rough and imperfect outline 
of Weismann’s theory, 1 proceed to apply it to the solution of 
some of the problems connected with variation in our British 
Ferns. The problem of variation itself has been already dealt 
with, because no doubt the cause of variation in Ferns is the same 
as the cause of variation in other forms of vegetable and animal 
life. 
If we come to examine Fern varieties in detail, we shall find 
that nearly all of them are cases either of excess or of deficiency 
of development in some part or parts of the plant. According to 
Weismann, these correspond to local inequalities in the nutrition 
of the germ plasm in the parent Fern. One thing you will 
all have noticed, viz., that you may sow spores of a perfectly 
normal Fern generation after generation without getting any 
appreciable variation in the seedlings. But if you once get a°break 
— once get a variation, however small, from the normal type — 
and sow from that, you immediately get a number of fresh 
variations. Take Mr. Barnes’s crested dilatatas and montanas, for 
example. Mr. Barnes first of all found a natural break differing 
comparatively little from the type— what we should call an 
ordinary variety. Sowing this, the result was a number of varieties 
much superior to (i.e., more abnormal than) the parent; doubt- 
less along with them were a large number inferior to the parent, 
and probably some quite normal ones. Mr. Barnes, however, like 
the wise man he was, thre\v the bad ones away and kept only 
the good ones. In this way almost any slight natural variation, 
by repeated sowings and selections, may be developed to almost 
