298 



NATURE 



[November 12, 1914 



modified, diluted, and blended, even though it origin- 

 ally appeared sharply and suddenly through a muta- 

 tion in the race. 



Returning to the previous line of argument, it 

 should therefore be obvious that no theory of heredity, 

 either Mendelism or any other, is adequate to account 

 for the origin of characters, and Mendelians are re- 

 sfKjnsible for much confusion of thought in failing to 

 observe the fundamental distinction between the origin 

 of a character and its subsequent inheritance. It is 

 the origin of the character, the nature of the change 

 involved in its appearance, and the manner and causes 

 of its apparition, with which mutationists are 

 primarily concerned. 



It is, therefore, only after analysing many actual 

 cases of mutation, and determining the nature and 

 causes of the germinal changes involved in each, that 

 generalisations can be made regarding the nature and 

 significance of the mutation process. This has been 

 done most completely in Oinothera among plants, 

 and by the experiments of Morgan and his students 

 with Drosophila among animals. Without drawing 

 comparisons between the two series of observations 

 and experiments, which differ in many features, refer- 

 ence may be made to a few of the points of view to 

 which the work with Oenothera has led. 



-In the first place, as already pointed out, it has 

 become clear and indisputable that mutation is a 

 phenomenon of variability and not merely of inherit- 

 ance. The breeding experiments and the cytological 

 investigations have co-operated in proving this thesis, 

 and thus eliminating the swarm of Mendelian hvpo- 

 theses. In the second place, we now know that muta- 

 tion, at least in Oenothera, is a composite process. 

 Each mutation is in a new direction, and represents 

 a different kind of germinal change. It is quite 

 erroneous to suppose, as Heribert-Nilsson has done, 

 that the various mutations of Oenothera fall into plus 

 and minus series. Any such classification is of the 

 most superficial kind, and the cytological investiga- 

 tion of the various mutants has done more than any- 

 thing else to throw light upon the real nature of the 

 change involved in each case. 



It may be worth while to observe here that the 

 multifarious nature of mutations, representing as thev 

 do departures in many directions from the parent 

 stock, is in harmony with present conceptions of cer- 

 tain general features of animal phvlogenv, according 

 to which many lines of divergence appear to have 

 departed simultaneously from a common stock. On 

 the other hand, orthogenetic phvlogenies, of which 

 there appear to be many examples in the palaeonto- 

 logical record, must depend upon other principles 

 which are not here considered. 



When we analyse cytologically the different mutants 

 of Oenothera they are seen to fall into at least four 

 or five categories. We may mention : — 



1. Duplication of a single chromosome to give 15. 



2. Hypothetically, in a case soon to be published, a 

 second duplication, to give 16 chromosomes. 



3. Triplication of the whole gametophyte series of 

 chromosomes, giving 21. 



4. Duplication of the whole sporophyte series, giving 

 28 chromosomes. 



5. The large class of mutations in which no change 

 in chromosome number occurs. 



The latter will be referred to again later. They 

 reallv include many diverse types of change, and only 

 agree ifi not having a visible structural change in the 

 nucleus. 



The peculiarities of the mutants lata and semilata 

 are constantly associated with the presence of fifteen 

 chromosomes instead of fourteen. In 1908 the meiotic 

 irregularit}- which leads to the formation of a germ 

 cell, and later an individual, having an extra chromo- 

 NO. 2350, VOL. 94] 



some was first observed. The same unequal chromo- 

 some distribution has since been observed in several 

 derivatives of Q£. himarckiana, and also in wild (E. 

 biennis from Woods Hole, showing that these are all 

 capable of producing lata mutations. It seems 

 obvious that the mutation occurs when a chromosome 

 passes into the wrong germ cell in meiosis, rmd that 

 the peculiarities in foliage and habit exhibited by lata 

 and semilata result from the fact that all the nuclei 

 of the plant contain a complex of fifteen instead of 

 fourteen chromosomes. The extra chromosome is, 

 evidently from its origin, a triplicate of a pair already . 

 present in other forms, and apparently any member 

 of the seven pairs may become the extra chromosome. 

 The obvious resemblances and differences between this 

 and the accessory or sex chromosomes need not be 

 considered here. 



The phenomenon of tetraploidy, now known to be 

 exhibited by many species of plants and animals, first 

 derived evolutionary significance from the fact that 

 CE. mut. gigas, which de Vries found to originate; 

 suddenly from CE. lamarckiana, is tetraploid, having' 

 twenty-eight chromosomes. The tetraploid species of 

 Hieracium, Antennaria, Drosera, Viola, Gyrostachys, ' 

 Potentilla, and many other genera have obviously- 

 originated in a similar manner, and in the phylogeny 

 of some families this doubling has occurred two or 

 three times successively, so that the process is one 

 of much evolutionary interest. The precise manner in 

 which the doubling of the chromosome series occurs 

 is not yet entirely clear, but it is obviously a different 

 process from the duplication of one chromosome which 

 gives rise to lata. 



In the majority of mutations there is no obvious 

 morphological change in the cell, and it appears that 

 in such cases the ultimate change is chemical. These 

 mutations fall into several categories differing in their 

 hereditary behaviour. Although this matter is too 

 involved to consider in detail here, reference may be 

 made to what appears to be the probable basis of 

 mutations which exhibit simple Mendelian behaviour 

 when crossed with their parent race. We may rake 

 as an example the case of CE. mut. rubricalyx, vshich 

 )riginated from rubrinervis in cultures in 1907, and 

 has never occurred before or since so far as known. 

 The original mutant was heterozygous, its offspring 

 vielding rubricalyx and rubrinervis in the ratio of 3 : i. 

 A homozygous race has since been obtained. 



The origin of such a mutant can be adequately 

 explained on the assumption that the substance of a 

 single chromosome underwent a chemical chanije of 

 such character that the activities of the chromosome 

 in the whole complex of the cell led to the production 

 of quantities of anthocyanin where it had not been 

 produced before or where it had been produced only in 

 small amounts. Of course, this change may be con 

 fined to one portion of a chromosome, and does not 

 necessarilv affect the whole body of the chromosome. 

 The behaviour of mutations which exhibit a simple 

 dominant or recessive unit-difference is fully accounted 

 for bv this hypothesis. It may be said that this 

 merelv assigns the phenomenon to the chromosome: 

 without further analysing it. But such hypothetical 

 changes in chromosomes are strictly analogous to th( 

 mutations which many bacteria are now known to 

 undergo, and if such simple organisms as bacteria can 

 experience sudden changes in structure and function 

 it is in no way improbable that chromosomes can do 

 likewise. 



We may even explain in this way the origin of 

 "duplicate' genes," such as have been described by 

 Nilsson-Ehle in wheat. In some families he obtained 

 a ratio of approximately 15 : i for red, and absence 

 of red in the glumes; iti other families the ratio 3 : i 

 was obtained. It appears that in the former case two 



