The Origin of Polyploids 471 



in a triploid or a tetraploid, a hexaploid or an octoploid results. 

 If a tetraploid plant becomes established among a group of di- 

 ploids it will probably produce further tetraploids by selfing but 

 it may produce some triploids by crosses between it and the 

 diploids. An apparent situation of that sort has been found in a 

 colony of Tradescantias in southeastern Louisiana. A railroad 

 track ran alongside a woods. In the woods grew the native T. 

 paludosa, a diploid species, but on the railroad embankment 

 running along the track were a number of tetraploid hybrids 

 between T. hirsutiflora and T. canaliculata, seeds of which had 

 apparently been introduced from more northern regions by the 

 railroad. Intermingled with them were a few plants of T. palu- 

 dosa. Three triploids were found near the region where the two 

 species met. They were undoubtedly hybrids between the native 

 diploid T. paludosa and the introduced tetraploids for they 

 showed characters of all three species. 



Division of the nucleus unaccompanied by the cell division 

 which almost invariably follows it can be brought about arti- 

 ficially by certain rather drastic changes in the environment of 

 the cell. Among the agents used have been sudden changes in 

 temperature, wounding with the formation of callus tissue, nar- 

 cotics, various other chemicals, bacteria, insects and similar in- 

 fective agents, changes in osmotic pressure, and radiation. Sax, 

 for example, subjected plants of Tradescantia paludosa to tem- 

 perature changes by keeping them at 8° C for two weeks and 

 then transferring them to a chamber kept at 38° C. Many chro- 

 mosomal aberrations were produced by this change resembling 

 those produced where the plants are exposed to X-rays. A num- 

 ber of different kinds of aberrations are produced, and among 

 them are found diploid pollen grains. A day after the removal 

 of the plants to the hot chamber many first anaphase figures 

 were found in which the spindle apparatus had been so dis- 

 turbed that the chromosomes fail to pass to the poles and 

 all apparently became included in one nucleus. After the plants 

 had been subjected to the high temperatures for four or five 

 days, complete asynapsis or failure of pairing was found in the 

 microsporocytes, there was no evidence of a spindle, and usually 

 all twelve univalent chromosomes passed into the resting stage 

 in a single nucleus. These meiotic abnormalities produce diploid 

 pollen grains, some of which apparently function and therefore 



