626 ON PHYSIOLOGICAL CHROMOMERES. 



and sesquiplex, having only one kind of functional pollen, but two 

 kinds of female gametes. Aberrant mutations may also occur and 

 sometimes in large numbers, but they are so rare that it seems 

 allowed to consider them as recombinations of the factors of the 

 main classes. 



To the lata-group we bring the semi-lata, delata, flava, subovata 

 and others, while the category of scintillans must include 0. mut. 

 nitens, distans and auricula which have yielded constant races, 

 besides a number of other forms, not yet sufficiently studied. 



On the basis of this view the production of 0. albida from 0. 

 lata, and that of 0. oblonga from 0. scintillans may be considered 

 as a hybrid splitting of the mutated races and this would account 

 for the relatively high percentage figures in contrast to the ordinarily 

 low mutation coefficients of other forms. We must then assume 

 that in the mutations which produce 0. lata the latent or inactive 

 factors for albida become active as well as those for 0. lata itself. 

 Or in other words, that the external conditions which provoke the 

 first, induce the second also. If the lata characters dominate over 

 those of albida, these latter will not show in the original mutant. 

 But in synapsis they will separate and the fecundation cannot alter 

 the result, since the characters are not transmitted by the pollen. 

 If there were no linkages a splitting after the rule for dihybrids 

 might be anticipated, but probably there are still more causes for 

 deviation, the weakness of the albida-type and of the laeta-lata 

 combination being perhaps among the most efficient. Further experi- 

 ments will have to explain the obvious deficiencies in the splitting 

 figures. But since neither lata nor albida reproduce their type by 

 means of their pollen, the ordinary methods of crossing cannot be 

 applied here, and different principles will have to be worked out. 

 The same considerations would explain the production of 0. oblonga 

 from 0. scintillans. 



The third category of chromosomes of (Enothera Lamarckiana 

 embraces the four smaller ones. They are assumed to contain the 

 factors for four different races, viz. 0. cana, 0. pallescens, 0. liquida 

 and 0. spathulata, all of which are dimorphic or splitting forms, 

 like 0. lata and 0. scintillans. Aberrant forms are still very rare 

 in this class, 0. candicans being a constant form of 0. cana with 

 only one kind of functional pollen, with dense spikes and a 

 lower stature, and 0. lactuca constituting a dwarfish type of 

 0. pallescens. 

 The assumed distribution of the factors for the main mutant 



