organism, on the embryo, and, as we have good reason to believe, on both sexual elements before impregnation has been effected. * * * * * {271} CHAPTER XXIII. DIRECT AND DEFINITE ACTION OF THE EXTERNAL CONDITIONS OF LIFE. SLIGHT MODIFICATIONS IN PLANTS FROM THE DEFINITE ACTION OF CHANGED CONDITIONS IN SIZE, COLOUR, CHEMICAL PROPERTIES, AND IN THE STATE OF THE TISSUES--LOCAL DISEASES--CONSPICUOUS MODIFICATIONS FROM CHANGED CLIMATE OR FOOD, ETC.--PLUMAGE OF BIRDS AFFECTED BY PECULIAR NUTRIMENT, AND BY THE INOCULATION OF POISON--LAND-SHELLS--MODIFICATIONS OF ORGANIC BEINGS IN A STATE OF NATURE THROUGH THE DEFINITE ACTION OF EXTERNAL CONDITIONS--COMPARISON OF AMERICAN AND EUROPEAN TREES--GALLS--EFFECTS OF PARASITIC FUNGI--CONSIDERATIONS OPPOSED TO THE BELIEF IN THE POTENT INFLUENCE OF CHANGED EXTERNAL CONDITIONS--PARALLEL SERIES OF VARIETIES--AMOUNT OF VARIATION DOES NOT CORRESPOND WITH THE DEGREE OF CHANGE IN THE CONDITIONS--BUD-VARIATION--MONSTROSITIES PRODUCED BY UNNATURAL TREATMENT--SUMMARY. If we ask ourselves why this or that character has been modified under domestication, we are, in most cases lost in utter darkness. Many naturalists, especially of the French school, attribute every modification to the "monde ambiant," that is, to changed climate, with all its diversities of heat and cold, dampness and dryness, light and electricity, to the nature of the soil, and to varied kinds and amount of food. By the term definite action, as used in this chapter, I mean an action of such a nature that, when many individuals of the same variety are exposed during several generations to any change in their physical conditions of life, all, or nearly all the individuals, are modified in the same manner. A new sub-variety would thus be produced without the aid of selection. I do not include under the term of definite action the effects of habit or of the increased use and disuse of various organs. Modifications of this nature, no doubt, are definitely caused by the conditions to which the beings are subjected; but they depend much less on the nature of the conditions than on the laws of growth; hence they are included under a distinct head in the {272} following chapter. We know, however, far too little of the causes and laws of variation to make a sound classification. The direct action of the conditions of life, whether leading to definite or indefinite results, is a totally distinct consideration from the effects of natural selection; for natural selection depends on the survival under various and complex circumstances of the best-fitted individuals, but has no relation whatever to the primary cause of any modification of structure.