62 
THE PLANT WORLD. 
When we consider mutation two theories present themselves. Muta¬ 
tion is either a permanent condition of plants or else it occurs periodically. 
If the former then it has no beginning and is not subject to experimental 
investigation. If the latter, it had a beginning and this beginning “ an 
external cause,” and the phenomenon becomes a fit subject for experi¬ 
mentation. “ The amount of mutability and its possible directions may 
be assumed to adhere to internal causes.” The lack of harmony between 
the demands of the biologists and those of the physicists concerning the 
duration of life on the earth disappears in the light of the mutation 
theory. All the facts point to the conception of “ mutation periods pro¬ 
ducing swarms of species from time to time.” The intervals between 
two successive mutation periods may have been centuries or even thou¬ 
sands of years. 
The principle of unit-characters and of elementary species leads to the 
recognition of two kinds of variability, first that due to acquisition of 
new units or to loss of those already existing, second lesser variations, 
due to the degree of activity of the units themselves. The latter is fluc¬ 
tuating variability and always takes place about a mean between two ex¬ 
tremes. 
Again we must distinguish between individual and partial fluctuations. 
Internal conditions determine averages. Individual fluctuations are due 
to external causes whose influence begins with the fusion of the sex¬ 
ual cells and may be termed embryonic variability. After germination 
the external conditions act separately on every part of the young seed¬ 
ling, and partial variability begins. “ Individual variability in the veg¬ 
etable kingdom might be called seed-variation, as opposed to partial or 
bud-variation.” Multiplication of the extremes of fluctuations by seed is 
always counteracted by individual variability, and thus extremes are best 
perpetuated asexually, when “ improvement ” along the desired line may 
result. This is the origin of nearly all of our largest and more palatable 
fruits, of the varieties of potato, and improved sugar cane. Varieties 
propagated vegetatively have the added advantage of being exempt from 
regression to mediocrity. 
Vegetative propagation is restricted to perennial and woody plants. 
Annuals or biennials must usually be propagated by seed, and races thus 
improved are notably inconstant. The standard must be kept up by re¬ 
peated selection, but this very act may result in improving the average 
and ameliorating the race. 
Consideration of the methods of breeders emphasizes the difference 
between artificial and natural selection. In artificial selection breeding 
ability (centgener power) is the chief standard for amelioration. Breed¬ 
ers choose seed from the parents having best average yield. But it is 
not possible to conceive how nature can make use of this principle. In 
a general way there is an analogy between breeding selection and natural 
selection, but this analogy “ points to the importance of selection between 
elementary species, and the very subordinate role of intra-specific selec¬ 
tion in nature. It strongly supports our view of the origin of species by 
mutation instead on continuous selection.” C. S. G. 
