The use of seeds of the desirable form, either from the original or from any of its derivatives, might give play to all the complex activity of the splitting of hybrids and of the free play of fluctuating variability. During the course of study of the fluctuating variability of a species, by means of successive generations grown from purely fertilized seeds, the observer may be so fortunate as to encounter individuals which do not fit into his series by reason of the possession of some quality not visible in the greater number of the progeny, or by the lack of such qualities. Thus in a progeny of red-flowered plants one may he encountered which has white bloom, or an individual with laciniate leaves may come in a pure progeny the remainder of which has entire leaves,
or a wholly glabrous specimen may be in a hairy progeny, or an individual may depart from the progeny in all these particulars. In either case it is apparent that the variability here is not one of the modification of a character, but by the total accession or loss of a character, and the variability is therefore a discontinuous one. In the progenies in which such variants have been seen hitherto, they form a proportion never larger than six or eight per cent, of the whole number and generally in much less quantity. If such mutants or saltatory variants be found, they should be closely followed, as they may furnish the observer with facts of the greatest value. Care should be taken to secure purely fertilized seeds and a minute anatomical examination should be made of the entire plant, with special respect to the characters in which it appears to differ from the mass of the fluctuating variants around the type.
