294 
R. Ruggles Gates 
and rubrinervis, treating them both anatomically and chemically. 
He points out that the differences between Lamarckiana and 
rubrinervis, are analogous to those between the twin types Iceta and 
velutitui produced in such a cross as biennis x Lamarckiana', and 
he believes that Lamarckiana and rubrinervis are each under 
certain circumstances capable of giving rise to the other. This 
view is borne out by certain results of the reviewer. It is also 
of interest that several collections of wild seeds of O. muricata from 
widely sundered parts of Canada show a corresponding dimorphism 
(21), indicating that such a condition may be widespread in this 
section of the genus. 
Three cytological papers have been published by Davis (4,5, 6) 
which are in large part a confirmation of the earlier results of Gates 
(13a, etc.) It was found, however, that in O. grandijiora the pairing of 
the chromosomes in synapsis was closer than in either biennis or 
Lamarckiana, closed rings being formed by the chromosome pairs 
in diakinesis, while in the latter two species and their derivatives 
the homologous chromosomes are very loosely paired or not paired 
at all at that time. Hence it appears that the attraction which 
causes pairing is greater in O. grandiflora than in the other species. 
As I have pointed out, the loosely paired condition gives greater 
opportunity for irregularities, such as actually occur, in the 
distribution of chromosome pairs during meiosis. It is now certain 
that this process is concerned in the appearance of some of the 
mutations ( lata and semilata). 
In a detailed account of somatic mitoses in (Enothera, Gates 
(17) found the number of chromosomes in an individual to be 
constant, the rare exceptions in metaphase groups being explicable 
in entire accord with the belief in the genetic continuity of chromo¬ 
somes from cell to cell. Certain peculiar cases were found in the 
cells of the nucellus in 0 . lata, in which the chromosomes were 
closely paired in metaphase as though about to undergo a reduction- 
division, though the chromosomes retained their somatic shape. 
Several recent papers have dealt with the question of the 
origin 0. gigas. There have been two chief views on this subject. 
Gates, in 1909, in showing the increased size of the cells and nuclei 
in gigas as compared with Lamarckiana, indicated the probability 
that the chromosome-doubling to give twenty-eight occurred in the 
fertilized egg or the young embryo, through a suspended mitosis. 
It was further pointed out that in various wild species the Ax 
number of chromosomes had probably originated in the same way. 
Strasburger (30) entirely concurred in these views and extended 
them to cover many new cases of 4.r or tetraploid species. In a 
subsequent paper, Gates (20) has brought together a list of over 
thirty cases of tetraploidy, natural or experimental, in plants and 
animals. Tetraploidy is therefore a well recognized evolutionary 
condition, and many new cases will doubtless be found when the 
chromosome numbers of more plants are known. The gigantic 
character of the cells in O. gigas accounts for most, though 
probably not all, the peculiarities of this mutant. 
The manner of origin of the tetraploid condition becomes, 
therefore, a matter of much interest. In addition to the view 
expressed by Gates and afterwards by Strasburger, Stomps (28) and 
Miss Lutz (27) have suggested another method, namely that tetra- 
