But--and this is the second point--modern research is also showing that the various species of Rust-fungi have split up into different varieties or specialised races, according to the particular host plants they inhabit. For instance there are special varieties or races of the particular species known as _Puccinia graminis_, the wheat rust, each of which grows well on various kinds of grain and grasses but refuses to infect others. Thus, the variety which infects Wheat refuses to infect Barley or Oats, while that variety which grows on Rye will not take on Wheat and so forth. Now it is important to notice that these specialised races are indistinguishable one from another by their visible microscopic characters: they are all botanically of the species _Puccinia graminis_ which forms its æcida on the Barberry. We must therefore conclude that we have here the same phenomenon as that met with in culture-races of bacteria which, having been fed for several generations on media rich in proteids, refuse to grow on media rich in carbohydrates, or when attenuated races are developed by culture under special conditions. Now since such physiological races as I have described are by no means confined to _Puccinia_ but are also known in _Melampsora_, _Gymnosporangium_ and other fungi, we must conclude from this and from what we know of variation in plants and animals generally, that variation and adaptation are common among parasites, insects as well as fungi. These considerations will serve to show moreover that the question of breeding disease-proof varieties of our cultivated plants is complicated by the danger of our breeding at the same time adapted races of their pests. It appears at first sight extremely improbable that we should escape the danger by breeding from those specimens of our plants which have best survived a fungus epidemic. Still, it must not be forgotten that "hardy varieties," and races adapted to other exigencies of the non-living environment, have been bred by selection--and nevertheless this variable non-living environment is always with us. The matter is therefore simply and solely one of experiment, and the retort that a disease-resisting variety of any particular plant has not yet been raised is no more valid than the objection that a true blue primrose has not yet been obtained: whether the same remark can be made with regard to any hope of a _disease-proof_ plant may be another matter, but in any case it must be made more cautiously in the light of our present experience. NOTES TO CHAPTER XVIII. The reader will find more on this subject in Bailey's _Survival of the Unlike_ and the literature quoted in the notes to Chapter VIII. For varieties of Indian Wheats, etc., see Watt, _Agricultural Ledger_, Calcutta, 1895. For a discussion on so-called "Disease-proof Wheats" consult Eriksson & Henning, _Die Getreideroste_.