EVOLUTION
662
EVOLUTION
than individual differences; for they are discontinu-
ous and constant, and are therefore capable of explain-
ing the gaps between existing species and those of
palipontologj'. We use the term single variation
when, from among a large number of offspring, some
one particular individual stands out that differs from
the rest in one or more characteristics which it trans-
mits unchanged to posterity. It is said to be peculiar
to the single variations that they cannot be reduced to
crosses. If this is possible, we speak of " analytical
variations". Favourable conditions for the appear-
ance of single variations are altered environment, a
liberal sowing of seed, and excellent nourishment. It
is a remarkable fact that the fertility of single varia-
tions decreases considerably, and this the more so the
greater the deviation from the parents. Besides, the
newly produced forms are comparatively weak. This
weakness and inclination to sterility are facts which
must be carefully weighed when determining the
probable importance of single variations for specific
evolution. Besides, it is — to our knowledge — in no
case excluded that the suddenly arising form may be
traced back to former crossings. Probably the only
case which is quite generally interpreted to demon-
strate specific evolution experimentally is that of the
prunrose observed by de Vries. After many failures
with more than 100 species, de Vr;es, in ISSG, deter-
mined to cultivate the evening primrose (CEnothera
Lamarckiana), whose extraordinary fertility had at-
tracted his attention. He chose nine well-developed
specimens and transplanted them into the Botanical
Garden of Amsterdam. The cultivation was at first
continued through eight generations. In all he exam-
ined .50,000 plants, among which he discovered SOO
deviating specimens, which could be arranged in
seven different groups, as shown in the following
table:—
and darker yellow, or smaller and lighter, the fruit
longer or shorter, the outer skin rougher or smoother,
etc.
It may be conceded that the Oenothera has devel- oped constant forms corresponding to the so-called "small or elementary species". The question, how- ever, is, whether the forms are really new ones or whether they owe their origin to some unexpected original cross. In fact, if we are to suppose a previous cross, perhaps 0. Lamarckiana and 0. sublinearis, then the 0. Lamarckiana of Hilversum had contained the different variations in a latent form and through culti- vation gradually reverted by throwing off the different variations. At any rate, there cannot be any ques- tion of a progressive development, for the reason that none of the new forms shows the slightest progress in organization or even development of any kind advanc- ing in that direction.
(3) Crosses and Mendelian Segregations. Cross- breeding can in nature hardly be considered as a factor in the progressive development of species ; in particu- lar, forms of different degrees of organization do not cross, and if they did, all deviations would soon be equalized according to the laws of chance and proba- bility. All the greater seems to be the importance of the jlendelian segregations. It may be kno\\Ti to the reader that the famous experiments of the Abbot Mendel were carried on with seven different pairs of characters which he crossed with one another, and then, by letting the cross-breds self-fertilize, he contin- ued the cultivation of the plants through a series of generations. In the first generation it was found that the offspring exhibited without exception the charac- ter of one of the parents, that of the other parent not appearing at all. Mendel therefore called the former — the prevailing — character the "dominant" and the other the "recessive". In the following generation,
Generation
O. gigas
albida
oblonga
rubrinervis
Lamarckiana
nanella
lata
scintillans
I. 1S86-S7
-
-
9
—
—
-
II. 1888-89
15000
5
5
—
III. 1890-91
1
10000
3
3
—
IV. 189.5
1
15
176
8
■ 14000
60
73
1
V. 1S96
2.5
135
20
8000
49
142
6
VI. 1S97
11
29
3
1800
9
5
1
Vir. 1898
9
—
3000
li
—
VIII. 1899
5
1
1700
1
-
ThespecimenofO.ji'jas (1895) was self-fertilized and
yielded 4.50 O. gigas forms, among which there was
only one dwarf form, 0. gigas-nanella. The three fol-
lowing generations remained constant. O. albida was
a very sickly form, though it succeeded, thanks to reg-
ular attention, in breeding constant offspring. Among
the 0. oblonga descendants there w-as one specimen,
albida, and in a later generation one specimen of O.
rubrincrris. 0. rutirincrvis proved to be as fertile as
Lamarckiana, and yielded besides a new variation,
leptocnrpa. The offspring of O. nanella was constant,
though among the 1800 ilescendants of nanella in 189G
three specimens showed oblonga characteristics. 0.
lata was purely female; but, fertilized with pollen of
other variants, it yielded 15 to 20 per cent 0. lata
descendants. O. scintillans was not constant. Ac-
cording to de Vries' observations fsince 188C), new
forms also originatcil in nature, but they succumbed in
the struggle for existence. The differences between
the single forms relate to various parts and degrees of
development, though in .several they are very slight.
The plants become either stronger or weaker, with
broader or narrower leaves; the flowers become larger
which was produced by letting the cross-breils fertilize
themselves, the recessive character appeared and,
moreover, in a definite proportion. On an average
this proportion was 2.89 : 1 or 3 : 1. In the second
generation 75 per cent of the whole number of plants
exhibited the dominant character, and 25 per cent the
recessive. No interniei.liate form.s were observed in
any case. In the third generation the off.spring of the
recessives was constant and remained pure recessives,
but among the offspring of the dominants some
remained constant dominants, while others were hy-
brids. The average proportion of the constant domi-
nants (D) to variable cross-breds (DR) was as 1 : 2.
Thus, besides the 25 per cent of constant recessives
(R), there was also 25 per cent (one-third of 75 per
cent) constant dominants (D) and 50 per cent (two-
thirds of 75 per cent) variable crossbrcds (DR) or
1D+ 2DR4- IR. The same proportion resulted from
the following generations of the crossbreils, and since
1900 this has been confirmed by otlier investigators in
the case of other plants (e. g. maize) and al.so of ani-
mals (e. g. gray and white mice).
Mendel's rule of segregation, therefore, runs thus:
