ANDROLETHAL FACTORS IN OENOTHERA. 631 



the vegetative nuclei from 14 to 15. It is absent in 0. Lamarckiana 

 mut. simplex, which has only 14 chromosomes, but which otherwise 

 behaves like the normal specific mutations. These have been called, 

 by Blakeslee and Gates, trisomic mutants, since one of the "pairs" 

 of their chromosomes must consist of three rods instead of two. 

 In the factor complexes of the specific mutat'ons of (Enothera 

 Lamarckiana gynolethal units must also be present, at least in a 

 number of instances. The best known among these are the mutants 

 lata and scintillans. Here only one-half of the mutated ovules will 

 produce viable germs after normal fertilization, while the other half 

 are either totally absent among the offspring or represented only 

 in extremenly rare individuals. In crossing 0. lata or 0. scintillans 

 with different species, four kinds of hybrids would be expected, 

 but as a rule, only three of them are observed. In the central chromo- 

 some the characters of laeta are opposite to those of velutina, while 

 in respect to the lata and the scintillans chromosomes one-half of 

 the gametes will carry the mutated complex of factors and the 

 other only normal marks. This would produce four kinds of gametes 

 in the ovules, viz. laeta, laeta-lata, velutina, and velutina-lata, producing 

 in crosses four types of hybrids. But, as has already been stated, 

 only three of them are regularly observed. The laeta-lata gametes 

 are almost never in a viable condition, and the same holds good in 

 the case of the laeta-scintillans gametes. Moreover, whenever such 

 gametes produce viable individuals after fertilization, these are 

 usually weak and wholly sterile, as has been found to be the case 

 in control cultures. 



We may now come to a discussion of the mutants semigigas and 

 gigas. As is well known, the number of their chromosomes is 21 

 and 28 respectively, the rods being doubled either on the maternal 

 side only or in both parental halves of the vegetative nuclei. In 

 semigigas this condition is explained by the presence of androlethals 

 for all of the chromosomes. These would kill almost every grain of 

 the pollen, leaving only those in a viable condition in which the 

 specific mutations had previously disappeared. As a matter of fact 

 the pollen of semigigas is almost wholly sterile and the stray viable 

 grains have been found to produce in crosses with the parent species 

 only hybrids of normal types. 



In 0. Lamarckiana mut. gigas we must assume these androlethal 

 factors to be absent. This would explain the high degree of fertility 

 of its pollen, as well as the transmission of its special characters 

 to its pollen-hybrids. Moreover it would explain the extreme rarity 



