THE CROSS-BREEDING &C. OF PEAS AND OF HARDY FRUITS. 469 
valuable fruit we have to combine many good points. Take, for 
instance, an apple—(1) flavour; (2) size; (8) colour; (4) earliness or 
lateness ; (5) cropping qualities; (6) constitution. Chance may favour 
the hybridist and he may hit the mark, in combining four or five out of 
the six, but the one “ missing ”’ character condemns the product as useless. 
Therefore, as we have not any guide as to these characters being either 
dominant or recessive in their generation, we have to make very many 
crosses before we succeed in combining them allin one plant. In the 
whole of my experience in cross-breeding, I have never been able to 
exactly reproduce any one of our seedlings, in all its characters, by re- 
effecting the same cross. [or instance, ‘ Royal Sovereign’ strawberry was 
raised by crossing ‘Noble’ x ‘King of the Karlies,’ and although this 
cross has been effected many times since, we have never reproduced 
just ‘ Royal Sovereign.’ 
THE CROSS-FERTILISATION OF THE PRA. 
‘The artificial cross-fertilisation of the pea, like that of most other plants, 
is very easily effected. It is not the actual crossing that is expensive and 
laborious, but it is in the after part that the labour and expense accu- 
mulate in the sowing and selecting, and re-sowing and re-selecting, and 
afterwards thoroughly fixing the variety. 
In practice, the bloom is opened in the very young bud stage, emascu- 
lated, and pollinated in the usual way. We do not find it necessary to 
protect each individual bloom from insects if the pollen is applied in the 
early stages, as the pistil is already weli protected from outside inter- 
ference by the pollen grains applied. The pea, being what is termed a 
self-fertiliser, is almost, if not perfectly, immune from insect interference. 
Hence the practicability of growing side by side various types of peas, and 
securing a true stock on re-sowing. The late Mr. Laxton was one of the 
first experimentalists to prove this, and to point the fact out to Darwin. 
After having effected the various crosses desired, and harvested the pods, 
the seed is sown in the following spring. In the first generation no 
deviation from the dominant seed-parent is noticed or expected. The 
second season the produce is again sown, and here we look for the breaks 
desired. If, of the parents of the cross, one is tall and one dwarf, we 
find nearly all the produce will be tall, but a few dwarfs and semi-dwarfs 
are also noticed, and are selected from the others. The appearance of a 
very large proportion of tall peas and a few dwarfs is satisfactorily 
explained by Mendel’s theory, and how far this law applies to semi-dwarfs 
which are also found (that is, intermediate in height to the two parents) 
I am at present unable to follow or explain. The produce of this cross is 
again sown the following season. All the dwarfs, semi-dwarfs, and talls 
being sown separately, we find in this generation a further splitting-up of the 
types, dwarfs still appearing amongst the tall selections in the proportions 
of the second generation, but the dwarfs and semi-dwarfs do not further 
break away in height, but come true. Therefore, we have to grow and select 
each cross through at least three generations before any fixed and definite 
type can be secured. Afterwards, the selected seedlings, if any—in many 
cases they are all useless—are further sown and tested, and we find “ rogues,” 
or false peas, still appearing for many generations, mostly of a very wild or 
