CASTRATION AND HYBRIDISATION IN THE GENUS HIERACIUM, 287 
is said to grow wild in Galicia. My reason for mentioning this species 
especially is that it is purely female, just as some Turauracwm forms. 
The anthers are present, but do not contain any pollen at all. This 
female form is able to set fruits when isolated under a coyer-glass—as 
indeed also after castration—and is consequently apogamic. Lut if we 
examine the fruits of a head, we find that besides the full ones not a 
few are empty (20 to 30 per cent.). Also in other species of Pilosellu 
such empty (barren) fruits occur among the fertile ones, while in the 
Archieracia nearly all the fruits are full. A probable supposition, which 
is supported by the cytological researches of Dr. Rosenberg, is the 
following: the fruits which under isolation remain empty need fertilisa- 
tion for their development; the developed fruits have been produced 
apogamically. The species being female, the first category of fruits is 
never developed unless there are specimens with pollen (hermaphrodite or 
male) in the native country of the plant, or unless the pollen of other 
species is able to fertilise them. The last case has been tried and with 
positive results, and this hybridisation confirms our supposition. 
I have produced several hybrids by crossing the H. exceliens with 
H, Pilosella and H. aurantiacum as fathers; the parents are the same 
species and the same individuals which afterwards eave apogamic fruits 
when castrated. The offspring of such a cross consist of a few hybrids 
and of many plants lke the mother; the cross-fertilisation has con- 
sequently succeeded only in some cases, the majority being apogamice. 
Another interesting point is that the sister-hybrids are not like each 
other. This was pointed. out by Mendel, since it formed a contrast to 
his results with Piswm hybrids, where the primary hybrids are all alike. 
The hybrids between H. excellens and H. aurantiacum have a deeper 
yellow flower-colour, and the under side of the outer corollas always bears 
a red stripe, which never occurs in H. excellens. The hybrid which is 
nearer to H. awrantiacum is hermaphrodite, the others are female like 
their mother; the latter are the stronger, especially with regard to their 
vegetative power. The characteristics of the hybrids between H. excellens 
and H. Pilosella are analogous ; the most important ones are the largeness 
of the flower-heads and the ramification of the scapes. 
The fructification-power of the hybrids is very limited; most of the 
fruits are empty, but the hybrids most nearly resembling H. excellens cive 
fuller fruits than the others. It seems then as if the influence of the 
mother is both the stronger and the better one, and herewith the fact is 
consistent that most of the hybrids produced are nearer to FH. excellens, 
a few are intermediate, and none quite like H. aurantiacum. 
After isolation or castration, the hybrids haye given some fruits which 
have germinated this year and wili soon flower. Here we have the 
interesting fact that the power of the parents to develop fruits without 
fertilisation has been inherited by the hybrids. 
At present I am not able to say anything about the second generation 
of the hybrids, as my sowings of last year, by some mistake, have failed. 
Mendel says that F, is quite like the first generation, and this seems 
very probable now, when we know that apogamy occurs in the genus. 
In another hybrid I have found some phenomena which do not 
agree with that, but which perhaps are to be explained in. another way. 
