376 
ON CROSSES AND 
different cells. Experiments should be made to ascertain 
whether, in cases of partial and imperfect fecundation, the 
pollen of another species, and even of nearly allied genus 
which could not alone fertilize the ovary, can act in conjunc- 
tion with a single grain, or at least with an insufficient quan- 
tity of the natural dust to effect the fertilization, and occasion 
the seed to produce a variety, not actually hybrid, but in 
some degree departing from the natural form. See, above, 
the account of Hymenocallis amoena, var. lorata, p. 211. It 
is certain, by the result of many experiments, made at Spof- 
forth, that the pollen of a nearly allied genus, which cannot 
effect fhe production of seed that will vegetate, will often 
cause some of the ovules to swell to a large, and occa- 
sionally to a preposterous, size, and become seed-like masses 
without an embryo, and the same circumstance has been 
observed in Germany ; and, as it can act so far, I do not see 
the impossibility of its influencing the character of the pro- 
duce, where the access of natural pollen is insufficient; and 
it seems to me questionable whether some of the singular 
varieties which occur among vegetables may not have been 
so produced. 
A very singular occurrence in the history of cross-bred 
plants took place last year in the garden of my brother 
(Hon. Algernon Herbert), at Ickleton, in Cambridgeshire, 
which deserves the attention of naturalists. In 1834 he 
purchased a plant, grafted from a hybrid Cytisus, known to 
have been raised in France between C. laburnum and pur- 
pureus, of which the leaves are as large as those of la- 
burnum, though a little different in form, the flowers of a 
dingy and rather coppery purple in long racemes. The 
plant purchased consisted of a strong laburnum stock about 
8 inches, and a grafted scion about 16, high. Its growth 
that year was vigorous. In 1835, from a strong branch, 
which was not in existence when the plant was purchased, 
proceeded a small shoot a foot and half long, covered 
with small leaves of the exact size and very nearly of the 
form of those of the little C. purpureus, while the rest of the 
tree, having reached the height of 8 feet, had the usual 
large foliage, approaching in appearance to that of the la- 
burnum. This extraordinary branch, which has in a great 
measure, though not exactly, assimilated itself to the struc- 
ture and habits of that one of the two parents to which the 
mule had originally the least resemblance, has this year, like 
