Vol,. 6, 1920 
GENETICS: BABCOCK AND COLLINS 
673 
€ntiation were lacking or non-functioning. We know that in all pro- 
meristematic cells there must be the potentiality for development of a 
complete individual but that during development something determines 
what, when and where tissues and organs shall be developed. In these 
hybrids that force is not acting, in a normal way at least, and hence the 
cells develop the various tissues in a haphazard way and otherwise mis- 
behave. 
The opinion has been expressed a number of times that the different 
species of a single genus having different chromosome numbers may have 
been derived, the larger from the smaller number, by fragmentation or by 
duplication resulting from non-disjunction during synopsis. Rosenberg^ 
has recently worked out the specific chromosome number of 15 species of 
Crepis (published 19, 4 having been determined by other cytologists) . 
He expressed the belief that the species having four and five pairs of chromo- 
somes could have been formed by non-disjunction occurring during the 
reduction division. Bridges^ has shown that such behavior is possible, 
at least for the sex chromosomes in Drosophila, for he has found flies with 
ten chromosomes, the typical number for the species being eight. The 
increased number came about through secondary non-disjunction of the 
sex chromosomes in both parents. This fly, a female, was able to live and 
reproduce. These two cases are not, however, equivalent since the ten- 
chromosome Drosophila had the regular female content plus two Y 
chromosomes which are not known to influence the production of charac- 
ters except that absence of the Y chromosome causes sterility of the 
males. 
Whether a zygote containing a duplicated pair of chromosomes can go 
through development may depend entirely upon which pair is duplicated. 
From our knowledge of the chromosome numerical variations in Oeno- 
thera and Primula certainly it is possible for a plant to develop to ma- 
turity with one extra chromosome or with an entire duplicate set. 
The behavior of the seven-chromosome Crepis hybrid leads us to believe 
that there is not such a direct relationship between the two parent species 
as Rosenberg suggests. However, this in no way does violence to Rosen- 
berg's hypothesis, for C, tectorum may behave in quite a different way 
when crossed with some other three paired chromosome species of 
Crepis. 
It is possible that more light may soon be thrown upon this side of 
the problem inasmuch as there are now 36 species of Crepis with which 
trial crosses are being made to further test their chromosome relationships. 
1 Clausen, R. E., and Goodspeed, T. H., Proc. Nat. Acad. Set., 2, 1916 (240-44). 
2 Rosenberg, O., Arkiv.fdr Botanik., 15, 1918 (11). 
3 Bridges, C. B., Genetics, I, 1, 1916 (1-52; 107-63). 
