1915] DE VRIES—A MENDELIAN MUTANT 343 
of high stature, and 25 per cent normal high specimens. This may 
be considered as sufficient proof that the splitting took place after 
the law of MENDEL. 
TABLE IV 
C. DwarFs AMONG THE OFFSPRING OF O. gigas (A, no. 3) 
ee | eee | oe. | eee 
Pe ea 242 I 0.5 
wi MANN See Sie 276 ° 0.0 
ie a ieee eG ry I o.5 
Ao eae te es 237 39 16.0 
Bey Ger ae 238 52 22.0 
ERE a 8 oP 236 5° 21.0 
Vai ode neon 196 42 21.0 
Cab ee 81 25 31.0 
Cor ea. 269 59 22.0 
Ree etal. 265 57 21.0 
The dwarfs were counted in June and July, and the degree of 
development at this time corresponded with the photographs given 
in my Gruppenweise Artbildung, p. 316, figs. 115 and 116. At this 
period they are clearly distinct from the normal specimens and so 
there was no difficulty in counting them. In some specimens of 
O. gigas mut. nanella the number of chromosomes has been deter- 
mined and was found to be the same as in O. gigas itself (28), as 
was to be expected. Partly on account of this fact, partly in con- 
sequence of the nearer relationship, the fecundations did not experi- 
ence the difficulties which are connected with crosses between 
O. gigas and O. Lamarckiana mut. nanella. They succeeded fairly 
well and yielded, as we have seen, relatively large numbers of seeds. 
The Mendelian behavior of the production of dwarfs by means 
of mutation from O. gigas, moreover, may be proved in another way. 
If the mutant hybrids of this form are fertilized by the pollen of 
O. gigas nanella, the expectation will, of course, be the production 
of 50 per cent of tall specimens and 50 per cent of dwarfs. But, 
on account of the smaller viability of the latter, we should have to 
be content with somewhat smaller numbers. In 1913, therefore, 
I crossed some specimens of apparently normal O. gigas with the 
pollen of a constant race of O. gigas nanella, my culture being the 
third generation derived from a mutant of 1910 (Gruppenweise 
Artbildung, pp. 315-316). I was fortunate in choosing, among 
