1872 . ] 
ON CROSS-BREEDING PELARGONIUMS.-NO. III. 
51 
green-foliaged sort, more than one-half of the offspring will be more or less 
variegated, and will also be wanting in vigour, while half will resemble the 
mother. Reverse the order of things here, and see how few will be found to 
resemble the mother. The difference in vigour between the variegates and the 
robust mothers that are employed in breeding for variegates, is greater than 
between those for flower, consequently the effects upon the offspring are still 
more apparent. 
From this conviction of the influence of the male parent, it will be inferred 
that in breeding for improvement in size, in combination with form, in the 
scarlets, I should employ the finest-formed flower as pollen parent upon the 
largest and thickest-petaled seed parent I could obtain, irrespective of form. The 
result of crossing two equally fine-formed and fine-quality flowers I find not to be 
satisfactory ; the flowers of their offspring are smaller, and degenerated in 
quality, the texture of their petals less smooth, and with a tendency to crumple. 
In fact, the same rule seems to apply to flowers as to animals—that the cross¬ 
ing of opposites more frequently produces the best results. The same analogy 
seems to exist with regard to what is termed breeding in-and-in; for I find 
crossing varieties the offspring of the same parents, or the offspring back again 
upon their parents, or the parents upon their offspring, to be decidedly pernicious, 
which indicates the necessity for procuring every season plants and varieties 
distinct from those hitherto employed. 
In breeding for improvement in the nosegay class, or for what are termed 
hybrid nosegays (a misapplied term, for they are not hybrid, the nosegays and 
zonals being certainly of one species), I apply the pollen of the nosegay upon the 
stigma of a broad-petaled zonal. 
As regards the blending and production of novel colours, much has yet to be 
done and learnt. My notes indicate that the magenta shades are the result of 
the mixtures of pink or lilac and scarlets; the magentas and scarlets produce 
various hues of crimson and maroon, according to the depth of the scarlet; white 
and scarlet, salmons ; and strange to say, salmon will not unfrequently result from 
the crossing of two scarlets ; but the extent of the blend, and the variations that 
result from the mixture of colours, depend, as I have previously observed, upon the 
respective constitutions of the plants employed. Probably no blend at all would 
result if the pollen parent possessed both a decided colour, and the greatest vigour. 
I find that the production of a brilliant and novel colour is accompanied by a 
primitive form of petal, which causes the difficulty in obtaining the combination 
of novel colour and good form. There is one fortunate circumstance, viz., that a 
bad constitution also generally accompanies the novel colour, as well as the bad 
form of petal. 
As regards the difference in the functions that is stated to exist between the 
pollen of the long and short stamens, I have not made sufficiently accurate 
experiments to form a decided opinion. The short seem to me to possess 
diminished power of fertilization, but this may be merely owing to the fact, that 
d 2 
