280 REPORT—1905. 
unremunerative above a certain point. But an improved variety, with- 
out costing any more to grow, may increase the returns by 10 or 
20 per cent., in some cases may nearly double them. 
As regards the value of selection, Wood shows that the composition 
of the mangel, which has been selected solely for such external qualities 
as shape and habit, has remained stationary during the fifty years or so 
for which we possess any information ; while between 1860 and 1890 the 
sugar beet has had its sugar content raised from an average of 10:9 to 
15 per cent. by the steady selection of seed-mothers for their richness, 
The prospects of breeding new varieties of wheat, and particularly of 
securing improvements in such qualities as ‘strength,’ have been enor- 
mously improved within the last year or two through the investigations 
which have followed on the rediscovery of Mendel’s law of inheritance, 
Wheat as a normally self-fertilised plant is particularly suited to the 
investigation of Mendel’s law, and the work of Biffen shows that, with 
a few possible exceptions, the characters of the parent varieties are 
inherited strictly in accordance with the expectations derived from a 
consideration of that law. The great practical importance of this 
generalisation lies in the fact that it thus becomes possible to pick out 
with certainty fixed types in the third generation of the hybrids, whereas 
without the guidance of Mendel’s law and working by the old plan of 
selection, followed by continuous ‘ rogueing,’ it was impossible ever to 
secure a pure strain unless by chance an individual possessing pure 
recessive or pure dominant characters had been hit upon from the 
first. : 
Biffen’s work further indicates that the power of producing a glutinous 
grain, such as will lead to ‘strength’ in the flour, is a Mendelian 
character, following the same laws of inheritance as the bearded or 
beardless habit or the colour of the grain or chaff. Extreme strength 
shown in any particular wheat can then be picked out and combined with 
any other essential qualities, such as the yield and the character of the 
straw, which distinguish our present varieties of wheat. Of course the 
inheritance of a quality like strength, which is only relative between 
different varieties, cannot be traced with the sharpness with which such. 
characters as the long-awned bearded type can be followed ; still the 
variation that is, as it were, superimposed upon the ‘strength’ or ‘weak- 
ness’ representing the inherited Mendelian character is not sufficient 
to obliterate the evidence of inheritance aceording to the law. And, of 
course, this variation of individual seedlings in the ‘strong’ section 
above and below the degree of strength possessed by the parent, i.e., the 
inherited character, gives the plant-breeder his opportunity of improving 
such a quality at the same time as he is combining with it the other 
characteristics that are desired in the new varieties. Biffen’s work 
among the wheat hybrids touches also upon another point of special 
importance to South African farming, where the incidence of ‘rust’ forms 
the greatest obstacle to extensive and successful wheat-growing. The 
climatological conditions which make for a rust attack have not been 
worked out, as far as can be judged from the behaviour of English wheats 
in various seasons, together with the prevailing climates in countries 
where rust is specially prevalent ; a flush of growth in the spring 
followed by high temperatures will favour the disease, but South Africa, 
with its great vaviations in the amount and incidence of the rainfall and 
with its very different temperatures, affords a very good opportunity for 
