332 EVOLUTION, GENETICS, AND EUGENICS 



A difference of opinion has long existed as to whether 0. gigas 

 arose from the fusion of two diploid cells or from a suspended mitosis 

 with division of the chromosomes in the fertilized egg. I am still in- 

 clined to adhere to the latter view as more probable, with perhaps a 

 sudden lowering of temperature just after fertilization as the cause. 

 Although it is a fact that diploid germ cells do sometimes occur, 

 diploid pollen grains are not known to be viable in plants. 



The frequency of polyploidy in flowering plants, and also in other 

 groups of plants, is one of the surprises of recent years. Not only does 

 the condition occur in such cultivated plants as pineapples, bananas, 

 mulberries, wheat, oats, sugar-cane, dahlias, and tobacco, but also in 

 such genera of wild plants as the roses, maples, chrysanthemums, 

 Erigeron, Hieracium, Rumex, Rubus, Crataegus, Spiranthus, and a 

 number of others. The multiplication of chromosome sets runs as 

 high as 8« in Rosa and in Acer, and even as high as iow in certain 

 species of Rumex (80 chromosomes) and chrysanthemums (90 chromo- 

 somes). Such a widespread phenomenon must be of fundamental 

 significance in the evolution of the genera in which it occurs. The 

 higher degrees of polyploidy are probably often connected with hybrid- 

 ization, but there the higher chromosome numbers are usually accom- 

 panied by apogamous (asexual) reproduction, which renders constant 

 even forms with an unbalanced chromosome number. This is true 

 even of the triploid mulberries and Erigerons, etc. In every case of 

 polyploidy the higher numbers have not arisen gradually by the addi- 

 tion of single chromosomes, but one or more complete sets have been 

 added each time and the process is a mutation involving considerable 

 discontinuity. 



That still other kinds of chromosome change occur, is shown both 

 from experimental work and by comparison of the chromosomes of 

 related species. Thus transverse segmentation of all chromosomes 

 has taken place in Primula kewensis, and end-to-end fusion of certain 

 pairs has evidently occurred in some species of Drosophila. In the 

 Japanese violets there is some indication that the small number of 

 large chromosomes in certain species may have been derived by the 

 fusion of smaller chromosomes found in other species. A process sug- 

 gesting transverse fragmentation of certain pairs of long chromosomes 

 appears to have occurred in various genera of Liliaceae. Further 

 study will no doubt throw light on the nature of these processes. It 

 appears already that the passage from one genus to another has not 

 infrequently been marked by a visible change in the chromatin mor- 



