ed that the effect of a
cross is commonly, though not always, to introduce fresh vigour into the
offspring, though why this should be so we are quite at a loss to explain.
Continued close inbreeding, on the contrary, eventually leads to
deterioration, though, as in many self-fertilised plants, a considerable
number of generations may elapse before it shows itself in any marked
degree. The fine quality of many of the seedsman's choice varieties of
vegetables probably depends upon the fact that they had resulted from a
cross but a few generations back, and it is possible that they often oust
the older kinds not because they started as something intrinsically better,
but because the latter had gradually deteriorated through continuous
self-fertilisation. Most breeders are fully alive to the beneficial results
of a cross so far as vigour is concerned, but they often hesitate to embark
upon it owing to what was held {164} to be the inevitably lengthy and
laborious business of recovering the original variety and refixing it, even
if in the process it was not altogether lost. That danger Mendelism has
removed, and we now know that by working on these lines it is possible in
three or four generations to recover the original variety in a fixed state
with all the superadded vigour that follows from a cross.

Nor is the problem one that concerns self-fertilised plants only. Plants
that are reproduced asexually often appear to deteriorate after a few
generations unless a sexual generation is introduced. New varieties of
potato, for example, are frequently put upon the market, and their
excellent qualities give them a considerable vogue. Much is expected of
them, but time after time they deteriorate in a disappointing way and are
lost to sight. It is not improbable that we are here concerned with a case
in which the plants lose their vigour after a few asexual generations of
reproduction from tubers, and can only recover it with the stimulus that
results from the interpolation of a sexual generation. Unfortunately this
generally means that the variety is lost, for owing to the haphazard way in
which new kinds of potatoes are reproduced it is probable that most
cultivated varieties are complex heterozygotes. Were the potato plant
subjected to careful analysis and the various factors determined upon which
its variations depend, we should be in a position to remake continually any
good potato without {165} running the risk of losing it a