But we see that the difficulty of selection is increased in the case postulated above, because two ideals are to be worked up to, and they may conceivably be incompatible. Not necessarily so, however, for breeders have solved such problems before in obtaining early _and_ heavy cropping races of potatoes, wheat, etc., sweet _and_ large grapes, strawberries, etc., hardy _and_ brilliant flowers, and so forth. There is, however, another aspect of this question of variability in organisms in this connection to be considered. Ever since cultivation began man has probably been cultivating not only the crops he desires, but also the pests which infest them, and if variation of his chosen plants occurs--and no one will deny that--surely variation of the fungi and insects which live on them also takes place. That this is so can be demonstrated, though, since it is not part of my theme to go into the question of peculiarities of species and races of parasites, the subject must here be passed over with a few remarks only. Recent researches have shown not only that fungi vary immensely in form and morphological characters according to the amount and kind of food-materials put at their disposal, thus bringing the whole question of polymorphism into the domain of experimental physiology, but that their capacities for infection, spore formation, etc., are also capable of variation and are dependent on the quality and quantity of food supplies, water, as well as on the temperature, illumination, and other factors of the environment. This is true of parasites as well as of saprophytes. _Botrytis_ forms conidia only in darkness and in moist air. Klebahn found that a _Puccinia_ growing on _Digraphis_ infected _Polygonatum_ readily and completely, _Convallaria_ imperfectly, whereas if sown on _Majanthemum_ it only just infected the plant and then remained sterile, while it refused to infect _Paris_ at all. Magnus has shown that _Peronospora parasitica_ can only infect meristematic tissues, and that when it co-exists with _Cystopus_ on _Capsella_, as is usually the case, it enters the latter plant by infecting the gall-like pustules of hypertrophied tissue induced by that parasite. Numerous parasitic fungi can only penetrate particular parts of plants. For instance, the _Ustilago_ of wheat can only infect the young seedling, and grows for weeks as a barren mycelium, only becoming a dominant fungus in the endosperm. Numerous other examples could be given, but these suffice to show some of the ways in which the nature of the food substratum supplied by the host affects the fungus. It is obvious that if the nature of this food changes, the fungus is also affected, and no doubt this is the principal reason why Rust-fungi, for instance, vary so much in their vigour and reproductive power on different wheats and grasses, though the other factors of the environment must also be of influence on them as well as on the hosts.