THE EVOLUTION OF PLANTS. 
27 
more than one single form. It was raised by Miss E. R. Saunders in her 
garden, and she informs me that she could arrange her plants in two sets, 
the one consisting practically of S. squalidus, and the other of an inter- 
mediate. From this intermediate seed was saved, and I have raised from 
this seed about one thousand plants, every one of which is like every 
other, without any evidence of a tendency to vary. There is no doubt, I 
think, that seed gathered from this crop will produce precisely the same 
plant. 
Streptocarpus. — Vouched by Mons. Mottet. Mons. Mottet regards 
these hybrids as coming true from seed indefinitely. Mr. Watson writes : 
"I am disposed to believe that if we ceased to interfere with the breeding 
proclivities of these plants we should get back to S. Bexii pure and 
simple." Messrs. Jas. Veitch & Sons write: "As no insect life, of 
which the agency is available, exists in our houses, the plants if left 
to themselves are practically sterile, and hence reversion to >S'. Bexii or 
any other species cannot be proved. All our seed has been obtained by 
the artificial fertilisation of selected varieties. This much is certain : 
a white variety pollinated from a white variety produces white varieties ; 
similarly for red and magenta coloured flowers, usually with some 
improvement in the colour. There is thus here a certain degree of 
constancy." 
Verbena. — Reputed hybrids of F. chamcedrifolia, V. incisa, V. phlogi- 
flora, and V. teucrioidcs. Vouched by Mons. Mottet. 
Viola (F. cornuta x * Blue King Pansy '). Vouched by Dr. Stuart. Dr. 
Stuart kindly gives me an account of the production of his strain by 
crossing as above. He says : " There may be something in what you say 
about hybrids reverting to the type, but I have not found it so." 
Boses. — Mr. Geo. Paul writes me : — " In Roses the H.P. ' General 
Jacqueminot ' type after some generations virtually reproduced themselves. 
Witness a seedling from the 'General' named 'Maurice Bernardin.' 
Seedlings from it were sent out as new distinct kinds, in perfect good 
faith, which after being exhibited proved too much ahke to the parent. 
In fact, the type grows so reproductive of similar flowers that we gave 
up breeding from it in despair. 
Bailey quotes the following genera as producing hybrids, which come 
very true from seed: — Aquilegia, Begonia, Cereus, Cirsium, DiantJncs, 
Erica, Geum, Hieracium, Hippeastrum, Lamium, Lavatera, Lunariay 
Salix, Veronica. Hybrids of Erica and Salix, he says, produce a progeny 
as pure as a pure species. 
Messrs. Jas. Veitch & Sons conclude their letter with this remark : 
On the whole the few cases stated above seem to imply a certain amount 
of constancy." This applies, I believe, to cases in which the progeny of 
a hybrid was so nearly identical with the parent that all the plants could 
be distributed under the same name. It does not appear to have occurred 
to them that a hybrid type, however much it might vary, could ever lose 
its character. 
Mons. E. Lemoine writes that he has never observed the reversion 
of a hybrid to either of its parents. He believes that races of hybrid 
origin may be fixed just as races are that are not of hybrid origin. This 
appears to express the general opinion. 
