548 
Reginald Ruggles Gates 
The hypothesis that such new units originated in some rnyster- 
ious manner in the germ plasm of 0. Lamarckiana and finally gave 
rise suddenly to the new form, 0. gigas, seems to the writer an 
uuphilosophical and unnecessary assumption with no facts to Support 
it. An iuvestigation of the size changes of the cells following on the 
production of the tetraploid number of chromosomes, certainly throws 
light on many of the external changes in 0. gigcis , which were pre- 
viously obscure, and further indicates the failure of the premutation 
hypothesis of de Vries to be of lasting value as a rnethod of ex- 
planation. 
Summary. 
1. Oenotliera gigcis has double the number of chromosomes pre- 
sent in 0. Lamarckiana and the other mutants so far examined, i. e. 
28, with perhaps 29 in certain individuals. 
2. The reduction phenomena so far as observed are similar to 
those in the other mutants. The first reduction division separates 
whole chromosomes. The second separates in some cases the longi- 
tudinal halves of these, but there are indications that in certain cases 
the second division mav be transverse. 
3. Measurements show that the cells in 0. gigas are conspicuously 
larger than in 0. Lamarckiana. In the case of the epidermal layer 
of the petals this relationship is almost exactly 2:1, which is in 
exact accordance with Boveri’s law for size of cells in sea-urchin 
larvae. In other tissues, however, the relative cell volumes of the 
two forms are different, e. g., in the surface cells of the Stigma 
about 3 : 1 and in the pollen mother cells 1,50 : 1. Thus while the 
cells of 0. gigas are larger in every tissue examined, yet the per- 
centage of increase varies greatly in different tissues. The larger 
cells doubtless account for the greater stature of this mutant. The 
comparative volumes of the nuclei of the pollen mother cells during 
synapsis were about 2 : 1. 
It is also found that the cells of 0. gigas are not increased 
equally in all dimensions e. g., the epidermal cells of the anther of 
0. gigas show an increase in length of 72,8 % over 0. Lamarckiana , 
while the increase in widtli is only 28, 4^- This will account for 
the change in shape of some Organs. 
4. These two factors, (a) increase in nuclear and cell size conse- 
quent upon or coincideut with the doubling in the chromosome number, 
and (b) change in the relative dimensions of the cells in some cases, 
