July, 1922] 
CLELAND OENOTHERA FRANCISCANA 
407 
occasionally seen at metaphase, but disappears in time without having 
undergone division, and seems therefore to have no part to play in the 
development of the nucleoli in the daughter nuclei. The fact that during 
interkinesis several new nucleoli can be seen to develop simultaneously 
in contact with the chromosomes, and probably from them, makes it seem 
probable that the nucleoli arise de novo, and not from preexisting ones. 
The nucleolus seems, therefore, to be a body which stores material of 
some sort, probably chromatin, and which is formed anew in each nucleus. 
Genetical 
It has been very clearly pointed out by Davis (1915) that the chief 
source of doubt concerning the validity of many of the interpretations 
based on breeding experiments in the genus Oenothera lies in the ever- 
recurring suspicion that the species under consideration are not genetically 
pure. He has shown that the mere fact of breeding true is not a sufficient 
test of purity, in view of the large percentage of pollen and seed sterility 
often encountered in these plants, which phenomena introduce an element 
of doubt, in that they allow of the possibility of the presence of more than 
one type of gamete, only one of which may survive to function, or, if they 
all function, to produce successful progeny. The element of doubt thus 
introduced into the study of even those species which seem for the most 
part to breed true tends to cast a shadow of suspicion upon the conclusions 
based upon this work. It becomes, therefore, a matter of great importance 
to develop methods by which the purity of a species may be satisfactorily 
tested, and to find species whose purity is beyond question. 
The main barrier in the way of determining whether the members 
of the Lamarckiana group are pure or impure is the high percentage of 
sterility there found. Until the reasons for this sterility are determined, 
it is impossible to tell whether "mutants" thrown by them are indeed 
mutants, or whether they are Mendelian segregates. Meanwhile, the 
abundant sterility is itself a possible indication of hybridity. There are, 
however, a few species of Oenothera in which the percentage of sterility 
is lower than in most of the Lamarckiana group, and which at the same 
time breed quite true. At least one of these species, however, Oe. biennis 
L., is not entirely above suspicion. While it breeds remarkably true in 
the main, it has nevertheless been shown by Stomps (1914) to throw a 
small percentage of "mutants." Furthermore, it shows over 50 percent pol- 
len sterility. The other species, however, which may prove to be the long- 
looked-for pure species, are Oe. grandiflora and Oe.franciscana. Both of these 
are extremely stable, and at the same time show a very small percentage 
of both pollen and seed sterilit3^ 
The latter species has proved absolutely stable thus far, as is shown 
by table i, published herewith through the kindness of Dr. Davis. 
