632 ANDROLETHAL FACTORS IN OENOTHERA. 



of its appearance as compared with the common production of speci- 

 mens of semigigas. The true gigas has only appeared once in my 

 cultures (1895) and from this individual the present race has been 

 derived. Semigigas is found regularly in cultures of a sufficient size, 

 amounting to one or more specimens among every thousand seedlings. 

 From the semigigas, of course, the gigas might arise by means of 

 a corresponding mutation. Such has been the case for a semigigas 

 found among the offspring of a cross simplex x biennis Chicago, 

 where an individual of the stout stature and the marks of the gigas 

 type with fertile pollen and 28 chromosomes has occurred in 1921, 

 constituting a new and beautiful race. But such occurences are 

 very rare, and the chance for 0. gigas to originate at once from the 

 parent type seems at least quite equal to that of an origin from 0. 

 semigigas. This latter is not, in its nuclei, halfway a gigas, but must 

 be considered as a gigas containing androlethal factors. 



From this discussion a line of useful experimental work may be 

 derived. It has often been desired to find a race of lata or scintillans 

 with 16, instead of 15 chromosomes, both rods of the mutated pair 

 being doubled. Such a race would be isogamic, having the same 

 characters in the pollen and in the ovules and transmitting its marks 

 by means of the male as well as by way of the female gametes. It 

 has never been observed. It must, however, be possible to produce 

 it by means of a loss of the androlethal factor of the old races of 

 the same name. The only way to reach this aim seems at present 

 to be to fertilize 0. lata or 0. scintillans with 0. gigas and to select 

 among the hybrids those which would combine the maternal type 

 with the desired number of chromosomes and constitute constant 

 races. Indications of the possibility of such a result are not wanting, 

 but they have not, as yet, been accompanied by cytological studies. 



Resuming the main lines of this discussion, we find that the tri- 

 somic condition, caused by non-disjunction of one of the chromosomes 

 combined with the presence of an androlethal factor, constitutes 

 the main difference between the specific mutants of (Enothera 

 Lamarckiana and the hereditary behavior of wild species. A possi- 

 bility of eliminating it and of returning to a uniform' and isogamic 

 condition without androlethals is indicated by the use of 0. gigas 

 in crosses with the specific mutants. In this way the difference 

 would disappear, at least mainly, and the comparison of experimental 

 mutants and wild species would lose one of its greatest present 

 difficulties. 



(Journal of general Physiology, Vol. 8, 1925. p. 109). 



