Gates, Tetraploid Mutants and Chromosome MechanismUs 145 



pollen grains exist, though it does not prove that they are func- 

 tional, b) An analysis of the facts in the bananas investigated by 

 Tischler, which are sterile, seems to require that a single tetra- 

 ploid cell developed a new individual and race by a mutation. 

 And the Primulas, studied by Miss Digby, in which a single "pin" 

 flower on one individual of the F l sterile hybrid between P. flori- 

 bmtda and P. verticillata gave rise in each of two independent 

 crosses to a tetraploid race, makes it seem probable that the bud 

 mutation which produced the "pin" flower was followed by normal 

 fertilization and doubling of the chromosome number in the young 

 embryo. In the case of the Mosses studied by the Marchals, it 

 is obvious that tetraploidy arose through the aposporous develop- 

 ment of a diploid gametophyte which afterwards produced diploid 

 gametes which by fertilization gave rise to a tetraploid sporophyte; 

 so that the essential change here was the production of a diploid 

 gametophyte. This is emphasized by the fact that in one moss 

 (Phascum ciispidatum} the apospory was accompanied by mutational 

 changes in the new gametophyte. 



3. The view which was held by myself, Strasburger, and 

 others, that 0. gigas and many other tetraploid species originated 

 through a suspended mitosis just before or just after the formation 

 of the egg, therefore remains to be disproved, and the facts seem 

 to require this explanation at least in some cases. In any case, 

 the evidence now at hand shows that in some plants the mutational 

 changes are not confined to the meiotic divisions but occur also, 

 1) in the aposporous development of a gametophyte (the moss 

 above-mentioned); 2) in bud mutations, such as the small-flowered 

 and small-leaved branch of an individual of 0. gigas Sweden des- 

 cribed in this paper; 3) probably in an early division of the egg, 

 in the cases of a periclinal and a sectorial chimera of Oenothera 

 referred to in this paper. 



4. In addition to the g/'gas of de Vries I have studied a race 

 which originated independently several years ago at the Botanical 

 Garden of Palermo, Italy, and whose characters are identical 

 with those of the Amsterdam giant. A third giant race originated 

 in the cultures of Nilss on at Lund, Sweden, from an independent 

 Swedish race of 0. Lamarckiana differing slightly from de Vries' 

 type. This giant is, as Nils son has shown, very distinct from 

 the gigas of de Vries. A narrow-leaved descendant of the 



of de Vries is also described in this paper. 



5. A preliminary statistical study of the pollen grains in these 

 giant races and in other Oenothera forms was made, and it was 

 found that offspring of gigas which differed somewhat in their 

 external characters differed still more strikingly in their pollen grains. 

 Thus in normal 0. g/'gas Italy there were about 28 43 per cent. 



XXXIII. 10 



