72 
R. Ruggles Gates, 
hypothesis of de Vries (1913) that triploid and tetraploid mutants 
arise only through the union of germ cells one or both of which is 
diploid is unnecessary, since both these conditions may arise in 
other ways. 
The period of fertilization is an equally critical time in the life 
cycle of an organism, and the evidence indicates that a number of 
germinal readjustments or mutations date from this period. Both 
triploidy and tetraploidy may perhaps arise at this time or shortly 
after. It has already been pointed out 1 that in crosses such as 
(E. nanella lamarckiana and its reciprocal, where both parent 
types split out in the Fj generation and both subsequently breed 
true, it is reasonable to conclude that some determining reaction 
occurs in the fertilized egg, in which one parental germ cell or the 
other gains the ascendancy. This was formerly explained by 
De Vries on the basis of pangens in different conditions. But 
since the important work of Muller (1918) on balanced lethal factors 
in Drosophila , it is evident that linkage to such lethal factors may 
explain this result, though more will need to be known concerning 
such lethal factors in (Enothera before the explanation can be 
applied in detail. Differential sterility depending on whether the 
lethal factor is linked to (i.e., is in the same chromosome with) the 
factor for dwarfness or that for tallness, would seem to meet the 
case. 
Concerning the origin of (E. lamarckiana mut. nanella , we may 
assume that it appears through the breaking of the linkage between 
the dwarfing factor and a lethal factor, for if it were a simple 
Mendelian recessive we should find occasionally a Lamarckiana 
plant which was heterozygous for dwarfness and gave 25% dwarf 
offspring. But such are never found in lamarckiana although, as 
we have already pointed out, they do occur in CE. gigas. This 
suggests that gigas has lost some of the lethal factors present in 
Lamarckiana. As a mutant, nanella has a frequency of only 1—2%, 
so that the cross-over between the dwarf and lethal factors must be 
an infrequent one. The discussion of lethal factors will be taken 
up again later. 
That mutations in other forms may date their origin from 
fertilization has also been held, and Punnett (1919) has recently 
concluded that this is the time of origin of the well-known cretin 
mutation in the sweet pea. This mutation differs prominently 
from the type in having a straight stigma protruding through a cleft 
1 The Mutation Factor, p. 222. 
