PLANT BREEDING AND HYBRIDIZATION. 209 



forms do not usually reproduce their kinds, but often result in rever- 

 sion to ancestral types, A number of examples were cited of reversions 

 which have taken place in the sweet pea, oiunt lavender prinudas, 

 Andalusian fowls, etc, which show that in a number of instances the 

 forms are not readily iixed, being the result of complex crosses that 

 are for the most part infertile. 



A paper b}' C. C. Hurst, entitled Notes on MendeFs Methods of 

 Plant Breeding, was read by the secretary, Mendel was apparently 

 the first to recognize the necessity of considering each single character 

 on its own mei'its. In selecting constant characters, he avoided con- 

 fusion by crossing only constant and hxed races of })lants, each of 

 which had been the product of repeated self-fertilization. If plants 

 are chosen for crossing, the ancestr}' of which is ludcnown, the result- 

 ing ojffspring will either be incomparalde or incomprehensible. The 

 writer cited his experiments with orchids in which by choosing con- 

 stant characters he had almost entirelj'^ succeeded in eliminating the 

 possibility of reversion. It is stated that some of the apparent excep- 

 tions to MendeFs results are probabh^ to be {ittributed to the crossing 

 of species which were not constant in cliaracter. The consideration of 

 differential characters was briefly discussed, and it was stated that 

 Mendel in his experiments always chose his characters in pairs, so that 

 they would be distinctly differential and capable of definite recognition 

 in the offspring. The more clear! 3' defined the differences between the 

 parental characters, the more marked will l)e the single characters in 

 the resulting offspring. The fourth point in MendeFs method is said 

 to b(^ distinctly new, and that is the crossing together onl}^ of dominant 

 and recessive characters. If one of the characters of the differential 

 pair is always distinctly dominant over tlie other, the latter is known 

 as the recessive character. Knowledge of this fact serves to give uni- 

 formity to the first generation and avoids the difficulty of continuing 

 through su1)sequent generations breeding in which the results secured 

 will not be uniform. The necessit}"^ of using large numbers of individ- 

 uals was pointed out. In this there was great advance over MendeFs 

 predecessors. After having secured hybrids they should be carried 

 through many generations. Mendel in all cases carried his experi- 

 ments to the third and fourth, and in some cases to the fifth and sixth 

 generations. Summing up the methods of Mendel, the author states 

 that hybridists who d(>sire to follow the footsteps of Mendel and help 

 elucidate the problems of inheritance will find it essential in tlieir 

 work to select parents possessing characters which are at once single 

 and constant, differential and dominant, and they must also take care 

 to raise large numbers of individuals through man}^ generations. 



A paper b}^ Hugo de Vries, director of the Botanical Gardens, Am- 

 sterdam, Holland, discussed Artificial Atavism, Atavism was defined 

 as the occasional restoration of an old type in a compound cross. 



