EXPERIMENTS IN HYBRIDISATION, &c. 121 
THe STABILITY OF HyBRIDs. 
Next to the question of the fertility of hybrids, perhaps the most im- 
portant to the practical breeder is the constancy or stability of hybrids 
fertilised with their own pollen. If, as is generally supposed, hybrids do 
not breed true to themselves, but. are inconstant and tend to revert to 
their grandparent species, then is the breeder’s work vain, and the econo- 
mic importance of hybrids considerably diminished. With a view to 
ascertain more precise facts than are at present on record in regard to this 
question, I have carried out some experiments in the genus Berberis, 
raising a large batch of B. x stenophylla from self-fertilised seed. 
Berberis x stenophylla is a primary hybrid between B. Darwinii and 
B.empetrifolia, originally raised by Messrs. Fisher & Holmes, of Hands- 
worth, Sheffield. 
The two parent species are very distinct. The one (B. Darwinii) has 
a vigorous and upright habit of growth, while the other (B, empetrifolia) i 
weak and drooping in habit. In the one the stems are thick, much 
branched, and covered with short, woolly, brown hairs; while the other 
has slender stems, little branched, and quite glabrous. In the one the 
spines are seven in number—short, spreading, and flat—while in the 
other they are three (one long and two short), set like an inverted T, each 
rounded and grooved below. In the one the leaves number three to four, 
broad, flat; with five to seven spiny teeth, shiny green above, lighter 
below ; while in the other the leaves are seven to nine, linear, closely 
revolute, erect, sharply mucronate, dull, dark green above, silvery below. 
In the one the flowers are racemose, six to twelve, orange-yellow, shaded 
with red without, pedicels rich red, flower segments long and narrow ; 
while in the other, the flowers are single or sub-umbellate golden yellow, 
on slender green pedicels, segments short and broad. 
The hybrid B. x stenophylla is intermediate in most characters 
between the two parent species. Its habit of growth is very 
vigorous, first upright and then drooping gracefully; stems me- 
dium thickness, branches long, pendent, slightly pubescent, with three 
spines of equal size. Leaves variable, four to nine, narrow, partly 
revolute, sub-erect, mucronate, dark green above, silvery green below. 
Flowers, sub-umbellate, one to six, deep yellow, pedicels reddish, seg- 
ments intermediate. Altogether the hybrid is fairly intermediate, though 
B. empetrifolia is prepotent in the number and habit of the spines ; in all, 
favouring that parent as 13 : 12, or 52 per cent. Five hundred seedlings 
of this hybrid were raised, and of these no less than 90 per cent. repro- 
duced the characters of the parent hybrid faithfully and well, with little 
variation ; while the remaining 10 per cent. varied from a form repre- 
senting about 66 per cent. of the grandparent B. empetrifolia, through a 
series of intermediate forms up to an extreme one, which reproduced 
about 67 per cent. of the characters of the grandparent B. Darwin. It 
is worthy of notice that not one plant out of the 500 completely reverted 
to either of the ancestral species, No. 18 being the nearest to B. Darwinii, 
and No. 1 the nearest to B. empetrifolia. 
The following condensed analysis of thirty of the most variable of 
these secondary hybrids will show at a glance how they differ from one 
