LOS, 
. 
DECEMBER 4, 1913] 
charts are drawn is not given in the text. The 
monthly isobaric charts published by the German 
South Polar Expedition, and partly discussed in 
detail, are not referred to. 
The daily weather charts are also utilised in 
determining the frequency of winds and storms 
in the different 10° zones of latitude; these are 
given in tabular form, but no conclusions are 
drawn from them. 
In the preface to the volume the president of 
the Royal Society, Sir Archibald Geikie, makes 
some very useful corrections to volume i. of the 
Meteorology of the National Antarctic Expedition. 
These refer specially to the question whether on 
the sledge journeys of the Discovery expedition 
the wind-directions were noted by true or magnetic 
bearings. W. Mernarvus. 
THE GROUP-ORIGIN OF SPECIES. 
Gruppenweise  Artbildung, unter  spezieller 
Beriicksichtigung der Gattung Oenothera. By 
Prof. Hugo de Vries. Pp. viiit+365+22 
coloured plates. Figs. 121. (Berlin: Gebrii- 
der Borntraeger, 1913.) Price 22 marks. 
T may be said at once that the facts in this 
volume represent perhaps the most compen- 
dious and extensive experimental treatment of 
hereditary phenomena which has yet been accom- 
plished in one limited group of organisms, and 
as such it deserves careful study by all students 
of genetics, The book is an outgrowth and further 
development of the views expressed by de Vries in 
“Die Mutationstheorie”’ (1901-03). Those views 
were, as is well known, founded chiefly upon the 
author’s experiments with Oenothera, the muta- 
_tion theory of sudden germinal changes being also 
based to some extent upon his conception of intra- 
cellular pangenesis. 
The present volume, therefore, marks not only 
an important advance in our knowledge of the 
hereditary behaviour in the ‘evening primroses, 
but also coordinates, and develops to a remarkable 
degree the views of the author on the general 
subject of heredity and its relation to mutation. 
The strength of this present work lies in the fact 
that the new empirical results all receive their in- 
terpretation in terms of the earlier theory. And it 
must be said that the enormous mass of experi- 
mental data with Oenothera has been coordinated 
and rendered intelligible in a striking way by the 
application of the author’s earlier conceptions. 
De Vries adheres to the view that characters 
which are independently inherited must be repre- 
sented by separate structures (pangens) in the 
cell, and one of the basic conceptions of the book 
is that these pangens are not simply present or 
NO. 2301, VOL. 92]. 
NATURE 
395 
absent from the cell, but may exist in one of three 
conditions: (1) active, (2) inactive, or (3) labile. 
On this basis the whole explanation, not only of 
several different types of hereditary behaviour in 
wild species and mutants, but also of the mutation 
phenomena themselves, is worked out. A theory 
which can bring into harmonious relation such a 
vast body of evidence is of much service, even 
though its validity may not be final. 
Since the phenomena of heredity occupy such 
an important part of the book, a few of the 
general results of crossing may be mentioned. By 
series of interspecific crosses it is shown that 
various wild species, including O. biennis L., 
O. muricata L., and O. cruciata Nutt., carry en- 
tirely different characters in their male and female 
germ cells. Such species are called heterogam- 
ous. The pollen grains usually carry a type corre- 
sponding nearly with the external characters of 
the species, while the egg cells may carry a very 
different type. Other species of Oenothera, such 
as O. Hookeri, O. strigosa, and O. Lamarckiana, 
are, like most wild species, isogamous, i.e., bear- 
ing the same qualities in their eggs and pollen 
grains. In heterogamous species the reciprocal 
crosses are, of course, unlike. 
In the subsequent crosses, several distinct types 
of hereditary behaviour are recognised, e.g. 
(1) twin hybrids—two types unlike either parent, 
and which subsequently breed true or split, being 
produced in the F,; (2) the formation of inter- 
mediate hybrids, which remain constant; (3) split- 
ting in F, into the two parent types, which after- 
wards breed true; (4) Mendelian splitting, in Fo. 
The mutations from O. Lamarckiana are thus 
classified according to the type of behaviour they 
exhibit. 
These types of behaviour again are discussed in 
terms of pangens. Why, for instance, does O. 
Lamarckiana x O. mut. nanella give dwarfs in the 
F,, while in O. mut. rubrinervis x O. mut. nanella 
dwarfs first appear in Fj? This is because the 
former cross represents a labile x inactive pangen, 
while in the latter we have the active x inactive 
condition. In this way many of the hereditary 
peculiarities of the Oenotheras are “explained ”’ 
by the same theory which explains the mutations 
themselves. It is considered that a mutation con- 
| sists in the change of a pangen from one condition 
to another, and sometimes in the formation of 
new pangens. These conceptions are largely in 
harmony with the cytological facts. 
Aside from these theoretical matters, one of the 
most important contributions of the work is to 
show by many instances that new and constant 
races frequently result from crossing—races the 
characters of which, moreover, are not Mendel- 
