106 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
seems to remain to us then one reasonable explanation, and that is 
parthenogenesis. : 
Parthenogenesis, or the production of fertile seeds in a capsule without 
fertilisation, is well known to occur in certain plants, e.g. Gnapha- 
lium, Mercurialis, and Ceelebogyne.* 
Professor Henslow, in his ‘“‘ Structure of Flowers,’’ p. 171, relates how 
Dr. Treub found a larva of an insect in the ovary of a Mauritian Orchid, 
Liparis latifolia, which seemed to feed on the juices secreted therein, 
without injuring either the ovules or the ovary. In a short time, without 
the aid of pollen, the ovules developed and became covered with seed- 
coats, as if under the influence of pollination; the irritation of the larva 
seemed to have developed the ovules in the same way the pollen 
tubes would have done. And this is possibly what has happened with 
these curious crosses. The pollen tubes of the distinct genera may have 
irritated and developed the ovules by feeding on the juices secreted by 
the ovary, yet by some incompatibility failed to fertilise the egg-cells, the 
result being seed-buds developed within the capsule, naturally bearing 
the characters of the seed-parent only. [Whether these seed-buds arise 
direct from the egg-cells, or from the antipodal cells as in cases of poly- 
embryony in Fuchsia, Allium, and Citrus, yet remains to be seen. T] 
There is, however, one slight difficulty in this explanation which 
puzzled me for some time, and that is that hitherto it was generally 
understood that in parthenogenesis all the plants were exactly 
alike from one capsule, reproducing the varietal characters of the seed- 
parent down to the minutest detail, just in the same way as ordinary buds 
and cuttings do; whereas, in the case of the Zygopetalum crosses par- 
ticularly, the seedlings varied slightly in colour, form, and size, both 
iter se, and from the seed-parent also. 
But recent experiments with “Daphnia” have shown that there is 
a certain amount of variation even in parthenogenetic offspring {; and as 
it has been already demonstrated that at least one nuclear division occurs 
in parthenogenesis,§ this might reasonably have been expected. The one 
slight difficulty, therefore, in accepting parthenogenesis as the explanation 
of the above curious crosses seems now to be removed. 
THE VARIATION OF PRimMARY Hyprips. 
Hybrids of the first generation between the same pair of species are 
found to have a certain specific likeness, yet at the same time they differ 
one from the other in varietal characters. 
Sex, per se, does not seem to have any influence in the variation of 
hybrids in the Orchidew (owing possibly to their being hermaphrodite by 
nature), the same varieties occurring both in the reverse and obverse 
crosses ; indeed, in several cases recorded, the progeny of the reverse 
cross and that of the original one have proved to be exactly the same. 
As we have seen in the inheritance of varietal characters, when a different 
variety is used as a parent, the result tends to be different, the variation 
generally corresponding with that of the variety used. But the variation 
* Kerner & Oliver, “ Nat. Hist. Plts.’’ ii. p. 469. 
+ N. H. P. ii. p. 469. 
t Roy. Soe., 4/5/99, “ Nature,’’ 59, p. 142, 
§ Weismann’s “ Germ Plasm,” 1893, p. 347. 
