STUDY OF SPONTANEOUS HYBRIDS IN EUROPEAN FLORA. 151 
Klotzsch propounded the principle that the sterility of hybrids is never 
due to imperfection in the ovary. But such a rule appears to me to be 
far from general. 
On the other hand, the presence of seeds only indicates a probability 
of fertility, so long as the seed has not been subjected to a germination 
producing positive results. 
When two very distinct species produce hybrids, one can divide the 
results into two classes: the one which favours one species, and the other 
which favours the second parent. These forms are usually intermediate, 
but without absolute fusion. Ina general sense one may say that the 
vegetative organs approach more nearly to one species and the reproductive 
organs rather resemble those of the other species, but yet without either 
being identical. 
The inversion of the 76/e of pollination has no apparent effect on the 
different series. 
In a nomenclature of spontaneous hybrids, there is no reason to 
indicate which is the pollen-bearing parent. It would be best to write 
A x B to indicate the series which is more like B, and B x A to represent 
the series more nearly allied to A. The form A x B, following the 
alphabetical order, has the advantage of by no means prejudging the 
hypothetical facts. 
The disjunction of the parental characteristics described in Naudin’s 
memoir is an exception. 
Floral anomalies in e¢ertain species are very frequent, but much less 
rare in hybrids than in the species. 
Certain families produce few or no hybrids. These are especially 
those in which the fertilisation is early, and has taken place almost as 
soon as, or even before, the opening of the corolla. Species with introrse 
anthers give equally few cross-bred products. 
Hybridisation between two species may be observed in one locality, 
but be more or less absent in another. This can easily be explained if 
the insects which carry the pollen are rare or do not exist in the latter 
region. 
Places where the foliage is so dense as to exclude the light are less 
visited by insects, and consequently have few hybrids. For this reason 
one place, which at one time might be favourable, might later be able, in 
the case of perennials, to preserve the hybrids for a series of years, and 
yet produce no more of them for a long time afterwards. 
For the same reason the clearings in woods are favourable places. 
For crossing, the flowers must occur at the same time, but it must 
not be forgotten that, in two species which do not usually flower simul- 
taneously, the earlier species may have late flowers, and the later one early 
flowers, which would bring about the necessary contemporary flowering. 
When the style is not as yet suitable for the anthers, the period of 
possible crossing is increased, and the probability consequently favoured. 
Another favourable circumstance is the existence of the species mixed 
together, or at all events in close contiguity with each other, and 
especially dicecism not merely of the species but of the genus. Bright 
colours and the presence of nectaries are also factors of importance. 
