REPORT OF THE CONFERENCE ON GENETICS, 93 
present and the former standpoint is well illustrated by taking two of 
the common ideas current among breeders and considering how each hag 
gained in precision. The ideas I shall speak of are those conveyed by 
the terms “pure-bred”’ and “reversion.’”’ We have at last a critical 
appreciation of the physiological meaning of the term “pure-bred”’ as 
applied to a plant or animal. In a general way every breeder is familiar 
with the notion that some animals and plants are pure while others are 
not. We have long been accustomed to distinguish the two conditions in 
various ways—estimating purity sometimes by truth to parental type, 
sometimes by the uniformity of the offspring. Neither of these tests, as 
we now know, is valid. An individual may be impure though not 
sensibly different from the accepted type of its breed; and though 
continued breeding from an impure individual will probably in the end 
reveal impurity, yet several generations may be produced in succession 
without any such indication appearing. For example, if in a rose-combed 
breed of fowls that had bred true for generations, a single-combed bird 
were to appear, we might formerly have supposed either that one of the 
parents was impure, or that a new variation had occurred. We now 
realise that the introduction of the single comb may have taken place in 
some generation indefinitely remote, and the appearance of that feature 
in a perceptible form is due simply to the fortuitous meeting of two germ- 
cells bearing the recessive character. 
An individual is pwre-bred when the two cells, male and female, from 
which it develops, are alike in composition, containing identical elements 
or characters. No long line of like progenitors is needed to produce a 
pure-bred plant. A purple sweet-pea may, as we now know, have been 
bred from white grandparents exclusively, and yet be pure to the purple 
character. Conversely, a white sweet-pea may be a seedling produced by 
the self-fertilisation of a purple-flowered plant, and yet be pure-bred in 
respect of whiteness. It matters not how the parents are bred. They 
may be mongrels, as heterogeneous in composition as packs of cards; but 
if from the two packs similar cards happen to be dealt, the product of 
these two cards is pure. And as in the cards we may consider” their 
attributes of colour, suit, and number as distinct, so in the living thing 
we know that the several features or physiological characteristics may be 
treated as distinct in the cell-divisions by which the germ-cells are formed. 
From this separability or distinctness of the characters it follows that 
an organism may be pure-bred in one respect and cross-bred in another. 
I need not remind my present audience that this conception of the unity 
and distinctness of characters provides the solid foundation which makes 
the science of Genetics possible. Instead of regarding genetic purity as a 
vague and problematical state which might or might not be attainable by 
a long course of selection and fixation, we now know exactly what it is 
and how it is produced. 
It is evident that this is a piece of knowledge which the practical 
breeder can turn to account. In future he will work with individuals of 
tested composition and avoid masses, thereby greatly simplifying the 
work of selection and fixation. It is no exaggeration to say that in this 
branch of industry the breeder can now perform in four years what 
formerly he could scarcely have effected in twelve. | 
