VEGETABLE HYBllIDITY. 89 
Linaria purpiireo-vulgaris, wlucli he had preserved for several years. 
la each case he had several hundred individuals. A good raany of the 
last generation partially or completely reverted to the L. mJnaris with 
yellow flowers, and a few to the L. purpurea. A greater number pre- 
sented no tendency towards either of the parent species, but neverthe- 
less did not resemble the hybrids of the iirst generation. They pre- 
sented all the phenomena of disordered variation. 
Similar facts occur daily in the practice of gardeners. The two cul- 
tivated species olPelunij (P. nyclaginlflora^ with white flowers, and P. 
violacea, with purple flowers) may be intercrossed and produce fertile 
hybrids. Those of the first generation are all alike ; in the second, 
they become remarkably diversified, and this variation increases until 
the plants are often monstrous, tlie changes being assisted by the arti- 
ficial impregnation of one variety by another. The same conditions of 
individual variability ai-e exhibited by a host of other cultivated flowers, 
of which M. Naudin cites especially the Primula and Eoses of our 
gardens. In like manner, as he indicates, the varieties of our fruit- 
trees are strictly individual in their nature, it being universally ad- 
mitted that it is only by grafting or budding that any particular va- 
riety can be propagated ; hence he concludes they also may be regarded 
as hybrids between several unknown specific types. 
But if hybridity, doubtless often produced by natural causes, such as 
the visits of insects to flowers of different but nearly allied species, be 
the cause of so much variation in cultivated plants, it becomes an im- 
portant question whether the same cause may not give rise to a similar 
effect in such as remain in the wild state. In some genera, such as 
the Salices, Potentillce^ Rumices^ etc, the intermediate forms between 
apparently w^ell-marked species are so numerous and so well graduated, 
that on examination it becomes difficult to limit the species, and these 
genera have always furnished subjects of dispute among botanists. 
In these forms the supposition that their numerous varieties maybe 
due to the influence of hybridization is rendered more probable by the 
fact that they present peculiar favourable conditions for intercrossing. 
Now if we suppose the crossing of two of these species to give rise to 
fertile hybrids which do not all revert to the parent types, disordered 
variability will come into play and produce, in a few generations, a per- 
feet chaos of undecided forms. 
The distinction between this disordered variability and the ordinary 
VOL. ill. [march 1, iSfi5.] H 
