22 
VARIATION 
if a sufficiently large number of individuals be measured, 
and a sufficiently small unit of measurement be used, 
the amount and nature of the variations can usually be 
expressed by a simple curve, rising from zero to a maximum 
and falling off again to zero on the other side. Thus the 
lengths of the lowest fruits of 568 plants of CEnothera 
lamarckiana were measured by De Vries with the following 
result, the upper figures being the lengths in mm., the 
lower the number of fruits of those lengths. 
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 
1 1 5 11 17 27 37 62 74 83 79 51 43 32 18 13 5 5 3 1 
Thus the greatest number of fruits have the mean 
length, and the figures are evenly grouped about this mean. 
Such variation is called continuous, and the curve is known 
as a binomial, Galtonian, or Newtonian curve, or curve 
of Frequency of Error. 
Frequently it happens that when a curve is plotted from 
a given character, it shows two or more humps or maxima, 
e.g. in a lot of Chrysanthemum segetum a counting of 
numbers of ray-florets gave maxima at 1 3 and 2 1 with a fall 
between. This is discontinuous variation, and indicates 
a mixture of two races, each with its own mean for the 
character in question. Seeds from n — 14-flowered forms 
gave a set of plants whose rays varied simply about 13. 
The important question now in dispute is whether this 
discontinuity arises from continuity by continuous infinitesi- 
mal stages, or by distinct steps. Such “step- variations,” 
in which one or more individuals appear with a variation 
not connected by continuous intermediate stages with the 
other individuals of the species, are termed sports (or, on 
one view, 7 nutations\ and if they be of a very pronounced 
nature may be monstrosities. Monstrosities graduate into 
sports by easy stages and it is difficult sometimes to decide 
under which head to class a given variation \ 
1 The study of monstrosities is termed teratology , and was formerly 
much employed in the decision of morphological problems. Thus the 
frequent occurrence of green leaves in place of carpels was regarded as 
a proof of the derivation of the latter from leaves, the doubling of a 
flower (i.e. the change of its stamens into petals) as a proof of the 
