At the very outset, and continued parallel with this course of breeding, a _Stud-book_ should be instituted, in order that all thus subjected to systematic improvement should be brought together, and thus initiate a general improvement of the breed. The development of these ideas will furnish matter for the following chapters. CHAPTER I. REGENERATION OF THE PERCHERON BREED. There are two ways of crossing applicable to any breed, both of which have had their earnest partisans. So much clamor has been made about them, I think, only because they have been simultaneously used and often mingled, and the results have been deranged by their use. This might have been avoided by commencing with the simplest and continuing with the best. The first may be called the renewal of a breed within itself, or interbreeding; the second, improving by foreign blood. We will pass them rapidly in review, trying to reach in the results the solid basis of truth. [Illustration: Decoration] CHAPTER II. REGENERATION OF THE BREED THROUGH ITSELF, OR BY SELECTION. The first manner, also called _selection_, consists in making, among the race itself, a rational, judicious choice of the most perfect types; those which are as free as possible from the most prominent defects of the breed; those which best recall the primitive type, if it possess the superior qualities which it is required to reproduce; those which, healthy and vigorous, seem to have among themselves the most affinity. This choice ought to be severe and rigorous, nor should we be discouraged by the small number of the elect. From the issue of this first selection, make a similar choice, and with them and their progeny march perseveringly in the same way, without ever looking to the right or to the left—that is to say, without ever listening to advice which would modify the work commenced, or to praises which might induce the desire for too rapid results. To proceed too fast is perhaps a still greater error than to stop on the way, inasmuch as it often renders a retrograde movement obligatory and reduces to nothing the results of several years of success. It is indispensable that the selections from which a good progeny is desired should be completely grown—that is to say, the horses should be at least four years past, and the mares fully three years old. Sell, without remorse, to the trade the least successful types, and most carefully keep the good. The horses, after serving some campaigns in their adult age, can be sold without inconvenience; a few well-proved types are sufficient for a district. But never part with the mares when they are remarkable for their conformation, temper, aptitude to work, and for their qualities as breeders.