THE MUTATION THEORY 493 



Thus in a race identified with 0. muricata the 14 chromosomes all form 

 a single circle; in 0. biennis they are usually in two interlocking circles, 

 one with 6 and the other with 8 chromosomes; while in the mutant 

 oblonga with 15 chromosomes the arrangement is more variable but 

 frequently shows a ring of 5 single chromosomes with 5 pairs attached 

 to it. In a trisomic mutant from 0. rubricalysXhewettii a varying 

 number of ring pairs was found by the writer. If the relative con- 

 stancy of these arrangements is confirmed, it will show an essentially 

 new type of integration in nuclear structure, that is, a fixed positional 

 arrangement of the chromosomes in the nucleus with relation to one 

 another. This wUl also confirm the hypothesis of the writer years ago, 

 that the homologous maternal and paternal chromosomes in Oenothera 

 are usually arranged alternately on the spireme thread before the 

 reduction division. The fact that also, as observed by the writer in 

 ipo8, pairs of chromosomes are frequently detached from the rest of 

 the spireme in certain forms at an early stage of diakinesis, would make 

 it easier for double non-disjunction to occur on the heterotypic spindle 

 in these forms. The further study of the chromosome arrangements 

 in hybrids and mutants will throw more light on these rearrangements, 

 and perhaps also on the nature of the forces which bring them about 

 Such rearrangements without change of number, if they prove to be 

 constant, are to be considered as mutations of another kind. 



Whether non-disjunction has played a part in the appearance of 

 species in nature with a different chromosome number is as yet un- 

 certain. But there are certain genera, such as Carex, in which the 

 haploid numbers usually increase by one from species to species. 

 Heilbom suggests that this may have happened by non-disjunction, as 

 in Oenothera lata, followed by the division of the extra chromosome to 

 form a pair. 



Polyploidy. — An increase in the chromosome number by multiples 

 of the haploid number (polyploidy) is a phenorhenon of considerable 

 phylogenetic significance in plants, although it appears to be relativel> 

 uncommon in animals. There has been a burst of new knowledge on 

 this subject m the last few years. Since the original mutant Oenothera 

 gigas has been shown to be tetraploid, and setnigigas mutations triploid, 

 we have an experimental basis for the interpretation of all such cases. 

 It appears probable that the triploid condition in Oenothera arises 

 through the union of two male nuclei with the egg, as this condition 

 has been actually observed in 0. lamarckiana by Ishikawa, and in other 

 plants as well. 



