REVIEWS. 125 
subject of discontinuity in the origin of species can scarcely be 
diminished. 
The section dealing with Œnothera is followed by one in which are 
recorded experiments upon selection combined with increased nutrition. 
The result supports De Vries’s contention that normal variability is a 
nutritive phenomenon, and that in most cases selection is simply the 
choice of the best nourished individuals. The author finds in several 
cases that high feeding and strict selection have a closely comparable 
effect. It would be a great advantage if experiments of this kind 
were to be repeated on a much larger scale and for a long series of 
generations. 
Section IV. is concerned mainly with experimental studies in the 
origin of garden varieties. These largely substantiate the view that 
selection does not lead to the production of specific characters. De V ries 
introduces certain new conceptions, which require a brief mention on 
account of their great interest for practical gardeners and breeders. 
They consist in the idea of races existing intermediate between a 
species and a complete variety or race of it. Such between-races are 
of two kinds, which do not usually co-exist ; moreover, either of them 
may occur even when the complete variety is quite unknown. In the 
case of a half-race a small percentage only of seedlings produce plants 
which show the racial character, and if seed is taken only from such 
plants this percentage does not notably increase. A mid-race , on the 
other hand, can be improved by selection, and normally either shows 
the racial character in about half its members or exhibits in the great 
majority of its members a combination of the character of the species 
with that of the race. For instance, an ordinary variegated plant is 
looked upon as showing a combination of green with the yellow 
character of a complete race—the aurea variety, which exists as such 
only in a few rare cases. On the other hand, many plants show a 
small proportion of variegated individuals at each sowing ; and this 
indicates the presence of the corresponding half-race. The relative 
development of the two co-existing characters is in such cases highly 
variable. 
It might be supposed possible to pass from the species to the 
half-race, thence to the mid-race, and so on to the complete race, 
simply by selection. De Vries shows that this is very rarely the case. 
The passage from a half- to a mid-race is a mutation (degressive), and is 
not more frequent than other mutations. 
As a further illustration of what is meant by a between-race, men¬ 
tion may be made of the five-leaved race of purple clover ( Trifolium 
pratense ) obtained by De Vries. It would appear that the plants 
occasionally found growing wild and bearing a single four-lobed leaf 
belong usually to a half-race. De Vries was lucky enough to find two 
plants upon which several leaves showed the anomaly, and from these 
by an elaborate process of selection extending over several years, a race 
