Gates, Tetraploid Mutants and Chromosome Mechanisms. 



less associated with greater seed-production (see table III). 7) They 

 were slightly slower in development, and five of them remained 

 rosettes. 8) As shown later, the flowers produced considerably 

 more pollen than yiyas Italy, though not so much as Lamarckiana. 



Fig. 6. 0. gigas de Vries, Amsterdam. Narrow-leaved Rosette. July 10. 



Cult. 180 / 12 . 



There is thus decidedly less sterility, both in anthers and ovaries, 

 than is the case with gigas Italy. 



The individuals of this culture showed a certain amount of 

 variation in crinkling (though in most the leaves were smooth or 

 nearly so), but No. I, 5 had constantly narrow leaves though differ- 

 ing from the type in no other respect. Another (II, 28) was very 

 distinct. It was late in developing but had broad leaves very much 

 crinkled with less toothed margins, rather closely resembling the 

 typical gigas of de Vries 25 ). 



understand that the chromosomes are being studied by Nilsson. By analogy I 

 should anticipate that the race will have at least 28, and possibly 30 or more 

 chromosomes 



25) Nilsson describes and figures this form (1912, p. 166, textfig. 27 left). 

 He attributes the differences between this individual and the type to the absence 

 of the factor for red leaf-midribs, which brings about correlatively all the other 

 changes in leaf and habit. It would be interesting to know how the numbers of 

 chromosomes in these two forms compare. 



