Page:Natural History Review (1861).djvu/148

136 casionally even necessary in plants apparently hermaphrodite. In other species, cross-breeding between individuals or races is rare and exceptional, and apt to be unfruitful.

Plants of distinct species breed together only under exceptional circumstances.

The hybrids thus produced are constitutionally more or less imperfect. They seldom produce a second generation, unless fertilized by an individual of one of the parent species, to which they then gradually return. They, therefore, do not establish permanent races, but disappear in nature, unless reproduced by a fresh cross-breeding between the parent species.

Setting aside, in the first instance, these hybrids, and accidentally abnormal extreme variations, monstrosities, and diseases, the variations of a species may be generally referred to two classes.

1. Variations resulting from the direct influence of soil, climate, food, or other external circumstances, such as luxuriance from a rich soil, fleshiness from a maritime exposure, &c. These act upon the individual; they may disappear in that individual when the exciting causes are removed, or they may become so engrafted on the constitution as to last through life, after removal of the causes; they may even become more or less hereditary through one or more generations. Seeds of a plethoric kitchen-garden vegetable, originally the result of a peculiar treatment in a rich soil, will, even under a different and uncongenial treatment, to a certain degree reproduce the same variety for some generations.

2. Variations which, arising from causes unknown to us, we consider as constitutional,—variations in the colour of the flower, in the form of particular parts, in the production or non-production of wings or other appendages to fruits, seeds, peduncles, &c. These, like variations in the features of animals, are often hereditary, and in plants under cultivation will last, or may be made (by selection of seed, &c.) to last almost indefinitely; and, in a wild state, they may, in particular localities, result in apparently permanent races. These races, however, generally breed readily with the typical forms of the species, and, although permanent and distinct in some localities, will generally, in some part of the area of the species, or under certain circumstances, be