28 Mutations and Evolution. 



The Swedish race of (E. lamarckiana also yielded not only 

 lata and semilata but also another form, having 15 chromosomes, 

 called incurvata. 1 One of the latter when selfed yielded 5 incurvata 

 and one like lamarckiana. In this case there is more probability 

 that a different chromosome was involved as the extra one, since 

 no latas appeared in the offspring of incurvata. But there is no 

 certainty owing to the small numbers involved, an unfortunate 

 result of the great amount of sterility in all these forms. 



Other forms having 15 chromosomes will be considered in the 

 discussion of parallel mutations, but it may here be pointed out 

 that a race of (E. biennis produces a lata mutation which is an exact 

 parallel to the lata from lamarckiana, and which has 15 

 chromosomes 2 . In the biennis from the Holland sand dunes, known 

 to have been naturalized there since the time of Linnaeus, Stomps 

 (1914) obtained a semigigas mutant. When pollinated by de Vries 

 (1915b) with pollen from pure (E. biennis without the flowers being 

 castrated, it produced 19 offspring, one dwarf, 10 biennis with 

 14 chromosomes and 8 of a new unnamed type having 15 chromo- 

 somes. The latter shows no resemblance to lata, but from the 

 description it appears to have several characters in common with 

 incurvata. It strongly supports the evidence from incurvata that 

 independent 15-chromosome types exist. Regarding its origin in 8 

 individuals from a cross, one must suppose that the reduction 

 divisions in the megaspores of (E. biennis semigigas produced 

 embryo-sacs, about half of which contained eight chromosomes in 

 their nuclei and half seven chromosomes, the remainder having 

 been extruded or lost through irregularities in division. That such 

 extrusion was not a haphazard and irregular phenomenon, however, 

 isindicated by thefactthat these 8 individuals all belonged to one type. 

 Another /a/a-Iike mutant was obtained by de Vries (1915b, 

 p. 186) from (E. lamarckiana x (E. biennis nanella. It had in 

 addition to the marks of lata the characters of the ordinary hybrid 

 type (E. lamarckiana x biennis, and confirms the conclusions of the 

 writer* regarding the origin of (E. lata rubricalyx and other /ata-like 

 forms in hybrids. 



Finally we may consider the important studies of Miss Lutz 

 (1916, 1917) on forms with an extra chromosome. It is to be hoped 

 that her counts of chromosomes, which appear hitherto to have 

 been made almost entirely from root-tips and other somatic tissues 



1 The Mutation Factor, p. 147. 



1 Gates and Thomas 1914. 



Gates, 1914, p. 265. 



