The order in which our subject will be treated in the present volume is as follows.  A long series of experiments will first be given in Chapters 2 to 6.  Tables will afterwards be appended, showing in a condensed form the relative heights, weights, and fertility of the offspring of the various crossed and self-fertilised species.  Another table exhibits the striking results from fertilising plants, which during several generations had either been self-fertilised or had been crossed with plants kept all the time under closely similar conditions, with pollen taken from plants of a distinct stock and which had been exposed to different conditions.  In the concluding chapters various related points and questions of general interest will be discussed.

Anyone not specially interested in the subject need not attempt to read all the details (marked []); though they possess, I think, some value, and cannot be all summarised.  But I would suggest to the reader to take as an example the experiments on Ipomoea in Chapter 2; to which may be added those on Digitalis, Origanum, Viola, or the common cabbage, as in all these cases the crossed plants are superior to the self-fertilised in a marked degree, but not in quite the same manner.  As instances of self-fertilised plants being equal or superior to the crossed, the experiments on Bartonia, Canna, and the common pea ought to be read; but in the last case, and probably in that of Canna, the want of any superiority in the crossed plants can be explained.

Species were selected for experiment belonging to widely distinct families, inhabiting various countries.  In some few cases several genera belonging to the same family were tried, and these are grouped together; but the families themselves have been arranged not in any natural order, but in that which was the most convenient for my purpose.  The experiments have been fully given, as the results appear to me of sufficient value to justify the details.  Plants bearing hermaphrodite flowers can be interbred more closely than is possible with bisexual animals, and are therefore well-fitted to throw light on the nature and extent of the good effects of crossing, and on the evil effects of close interbreeding or self-fertilisation.  The most important conclusion at which I have arrived is that the mere act of crossing by itself does no good.  The good depends on the individuals which are crossed differing slightly in constitution, owing to their progenitors having been subjected during several generations to slightly different conditions, or to what we call in our ignorance spontaneous variation.  This conclusion, as we shall hereafter see, is closely connected with various important physiological problems, such as the benefit derived from slight changes in the conditions of life, and this stands in the closest connection with life itself.  It throws light on the origin of the two sexes and on their separation or union in the same individual, and lastly on the whole subject of hybridism, which is one of the greatest obstacles to the general acceptance and progress of the great principle of evolution.