96 THE TUBEROUS BEGONIA. 



of fact, there is a great deal of chance work in hybridising, especially when 

 violent crosses (i.e , between widely-different individuals^ are made. Any one 

 cross, if of an experimental nature in particular, may be a great success or 

 may turn out a failure ; yet judicious selection, combined with perseverance, 

 are bound to tell in time, and will always ensure success in the end. 



If, on the other hand, any particular characteristic, such as depth or purity 

 of colour, great size, dwarf habit, or the like be desired, then both the plants 

 selected for parents should possess the required quality in a marked degree ; 

 and by again selecting those of the seedlings that most nearly approach the 

 desired standard, in time something approaching perfection, or at any rate a 

 very high degree of excellence, will be attained. 



This may appear something like a contradiction of what has been said pre- 

 viously, but though extreme crosses are undeniably useful in their way, yet 

 it does not do to put all our eg^s in one basket and to develop any one charac- 

 teristic, or to secure and bring to perfection any particular class or type of 

 flower that may have been obtained ; the surest, and, indeed, only way is to 

 breed solely from those that already exhibit the character or tendency required 

 in a marked degree. But the breeders should riot be related, if this can be 

 avoided, or if not, let the relationship be as distant as possible. For instance, 

 if two of the darkest- coloured red or crimson flowers in the collection are inter- 

 crossed, they will ultimately produce a much deeper shade, though two or 

 three, or perhaps several generations of seedlings may have to be raised before 

 the desired result is obtained. 



Again, the erect- flowering type, which is one of the most useful and effective 

 departures that have yet occurred, was obtained simply by inter-crossing 

 examples that showed a tendency to hold their blooms upright, and anyone 

 with a few dozen plants to select from may attain the sanu result in the 

 course of two or three years ; there is also plenty of room for improvement 

 in this class still. 



It will thus be seen that any flower possessing some good points, but 

 deficient in one or two respects, may have these faults amended perhaps 

 entirely eradicated in the course of a generation or two by being crossed with 

 another variety possessing the characteristics in which the first was wanting, 

 and at the same time any desirable "break" or style of flower may be secured 

 and brought to perfection by working on examples exhibiting a more or less 

 decided tendency to the desired type. 



After working among these flowers for two or three years, and having become 

 pretty well conversant with the characteristics of the different varieties, it is 

 surprising how many more or less distinct types are found to exist, defined not 

 so much by colour, though this to some extent is a guide, as by the form and 

 character of the blooms, by the habit of the plant, and some peculiarity of 

 foliage, etc. When thoroughly at home among them, one can almost always 

 tell with tolerable certainty the source from which any seedling possessing any 



