ON PHYSIOLOGICAL CHROMOMERES. 625 



of velutina. When combined to one another they yield the Lamar- 

 ckiana and this, as is easily seen, will repeat the splitting into laetu 

 and velutina gametes in every generation. The result will be a con- 

 stant race of Lamarckiana, with two kinds of sterile seeds. 



This peculiar constitution gives some likeness to hybrids of the 

 ordinary type and this fact has induced some authors to the very 

 unlikely conclusion that 0. Lamarckiana might really be of hybrid 

 origin. If this were so, it was hoped, that it would perhaps be possible 

 to consider the splitting phenomena of our plant not as real mutations 

 but only as the results of this special condition. But the two com- 

 ponents can never have existed as independent races, simply on 

 account of their lethals, which would have killed all of their seeds 

 from the very beginning. Therefore they can never have produced 

 a hybrid. Moreover the hybrid-like condition is of very common 

 occurrence among a large category of wild species of (Enothera and 

 it would be very arbitrary to assume for every one of them two 

 hypothetical parents. Such have never been discovered. 



On the other hand the hypothesis of a hybrid condition would 

 not help us in explaining the splitting phenomena of 0. Lamarckiana. 

 These are of quite a different nature. No hybrids have as yet been 

 discovered which might produce the same type of offspring. This 

 is essentially true for the typical mutations, which can in no way 

 be considered as recombinations of previously existing characters. 

 And as for the Drosophila-like mutations, the hypothesis would 

 evidently be of no use at all. 



Our second category of chromosomes included the two remaining 

 large ones, which we call lateral in order to have an easy designation. 

 It has already been pointed out that each of them must contain 

 at least two physiological chromomeres in a mutable condition. In 

 the one the two linkage groups determine the features of 0. lata 

 and 0. albida, in the other those of 0. scintillans and 0. oblonga. 



0. lata has weak stems, rounded leaves, cupshaped flowers with 

 crinkled petals, scanty pollen or no good pollen at all, handshaped 

 stigmas a. s. o. 



0. albida has a pale foliage with a soft pubescence, erect stems, 

 a low stature, small flowers .and fruits and is ordinarily biennial. 

 O. scintillans has dark-green and shiny leaves, a long spike of flower- 

 buds above the opening flowers, and small fruits. 0. oblonga has 

 broad and white midveins, a narrow spike and thick, rounded 

 fruits. Moreover 0. lata and 0. scintillans are dimorphic and split 

 in every generation, while 0. albida and 0. oblonga are constant 



40 



