45 
which it was impossible to increase the size of the grains. 
By this process of selection the investigator had probably 
succeeded in isolating a pure strain of the particular 
variety with which he was experimenting, characterised by 
a larger seed than the sample with which he started, which 
was, no doubt, a mixture of races, some producing smaller 
and some larger seeds. But once a pure line had been 
obtained, though it showed slight fluctuations, yet it could 
not be improved by selection of the extremes of these 
fiuctuating variations. 
We know, however, of other variations which arise from 
time to time in all species of plants, and which are very 
dilferent' from the fiuctuations as we may term those 
described above. Sometimes we come across new forms 
which differ considerably from the normal type in one or 
more characters, as, for example, the cut-leaved varieties of 
so many plants. In this case we do not find a series of 
forms intermediate between the cut-leaved individual and 
those with normal foliage. This second type of variation 
which arises suddenly Darwin calied a sport, and he con- 
sidered it to be of comparatively little importance in the 
evolution of plants, as its very infrequent occurrence would 
cause it to disapeede in nature by constant inter-crossing 
with the more numerous normal forms. By artificial selec- 
tion, however, man can perpetuate and establish these 
sports, as has obviously been the case with many of those 
forms which took the fancy of the horticulturist. Itisjtothis 
kind of variation that we owe so many of our interesting 
and peculiar forms of cultivated plants. Recently a Dutch 
botanist, De Vries, has endeavoured to show that this 
second kind of variation, which he calls muéatiun, to dis- 
tinguish it from the former or fluctuating variation, is of 
more frequent occurrence than had been supposed by 
Darwin. He observed that a certain large group of pianis 
of Evening Primrose which had established itself in a wild 
condition in Holland showed a very considerable amount 
of mutation, and his experiments and other observations 
led him to the conclusion that at certain periods, possibly 
owing to changed environment, plants passed into a phase 
of mutation, during which numerous new sub-species or 
races might arise. Interesting and important as De Vries’ 
experiments are, his case cannot be considered proved until 
we know more about the previous history of the plants 
which show such considerable mutation. At present we do 
