214 



NATURE 



[November 6, 1919 



the second is the way in which varieties are known 

 to arise in horticulture and arboriculture. 



Since the publication of my book on the muta- 

 tion theory (1901-3) numerous instances of muta- 

 tion have been observed by different investigators 

 among animals as well as among plants. Half a 

 dozen species of CEnothera, some types of 

 Primula, the walnut, the sunflower, Narcissus, 

 Antirrhinum, Ligustrum, and many other in- 

 stances might be cited. Among insects Morgan 

 and his pupils have described more than a hundred 

 mutations from the fruit-fly, Drosophila. Other 

 cases have been studied by Tower for Leptino- 

 tarsa, etc. 



The production of new races of agricultural 

 crops by means of continual selection constituted 

 for Darwin one of his strongest arguments. He 

 showed conclusively that new species and varieties 

 are produced in Nature in the same way as agri- 

 cultural novelties. But at that time the practical 

 method was far from being clearly understood. 

 The work of Hjalmar Nilsson and Hays has since 

 shown that selection may be conducted according 

 to the principle of the mutation theory, only one 

 choice being necessary to start the whole new 

 variety. 



It is now generally conceded by mutationists 

 that the initial change takes place in the produc- 

 tion of the sexual cells before fecundation. From 

 this conception it follows that the chance of two 

 similarly mutated cells to meet one another in 

 this process must be very small, whereas ordin- 

 arily the mutated cells will combine with normal 

 ones. This must produce half-mutants, and these 

 may, in ordinary cases at least, split off the full 

 mutants after the same rules which Mendel dis- 

 covered for his hybrids. Sometimes the half- 

 mutants will be distinct from their ancestors, as 

 in Oenothera Lamarckiana rubrinervis and 

 erythrina, and, therefore, will easily be discovered. 

 In other instances external differences may be 

 absent, and only the unexpected production of a 

 new type in about 20-25 pcr cent, or more of the 

 individuals will betray the internal change. This 

 explains the mass-mutations discovered by 

 Bartlett. Such an indirect way of producing 

 mutations by means of two successive steps seems 

 to be very common in Nature, and will probably 

 afterwards prove to be the general rule. 



Willis has made an elaborate statistical study 

 of the appearance of endemic species, which he 

 considers to be the youngest of their region. He 

 finds that utility of the new characters cannot 

 have had any part in their production, since it 



cannot be shown to have any influence either on 

 their first local extension or on their subsequent 

 spreading over larger regions. Wide spreading 

 is mainly the result of age, the oldest species 

 having, as a rule, the largest areas. Moreover, 

 in comparing the diagnoses of endemic species 

 with the differences among the mutated forms of 

 such a group as the evening primroses one finds 

 a close parallelism, showing that our experimental 

 mutations are quite analogous to the species- 

 producing steps of Nature. 



Objections against the mutation theory have 

 been njade by different investigators. Some 

 systematists and palaeontologists still adhere 

 to the old view either wholly or only for special 

 cases. Biologists rarely attack the theory in a 

 direct way, but mainly discuss the question 

 whether the observed mutations are really the 

 representatives of the species producing changes 

 in Nature, as is claimed. They assume that the 

 splittings seen in our experiments are due to 

 hybridism, and that every mutating species is a 

 hybrid between supposed ancestors which pos- 

 sessed the mutative characters as specific marks. 

 This idea can scarcely aid in simplifying the ques- 

 tion, since it puts the origin of the characters 

 involved on to unknown parents. Sterile varieties 

 cannot produce hybrids, and therefore cannot 

 originate in this way. This fact seems sufficient 

 to disprove the hypothesis. In the case of the 

 evening primroses this view has led to fantastic 

 diagnoses of hypothetical ancestors, but even 

 these fail to explain the facts observed in our 

 cultures. Morgan's hypothesis of crossing 

 over, which goes far to explain the splitting 

 phenomena of the fruit-fly, fails in its application 

 to the evening primroses, since here half-mutants 

 are the rule. These must evidently be produced 

 without the aid of that process. Moreover, the 

 heterogamous mutants have dominant characters 

 which are handed down by the egg-cells, and not 

 by the pollen, instances of which are given by the 

 mutations called lata, scintillans, cana, liquida, 

 and others of Oenothera Lamarckiana. Evidently 

 these can never be explained by the assumption 

 of a hybrid condition of the parent species. 



Thus we see that the broad arguments for the 

 mutation theory are continually increasing in 

 number, whereas the criticisms are more and more 

 directed against special' cases. They are concerned 

 with the possibility of experimental proof and with 

 the fitness of our material for further studies, 

 but are not expected to invalidate the theory as 

 such. 



THE PROGRESS 



By Prof. W. 



FROM the discoveries to which the Mendelian 

 clue immediately led, many lines of research 

 and speculation are diverging. These enterprises 

 have still aims in common, a fact which we 

 recognise by including all under the one name, 

 NO. 2610, VOL. 104] 



OF MENDELISM. 



Bateson, F.R.S. , 



genetics ; for, thougfh various in their methods, 

 I all relate to the physiology of breeding, a depart- 

 i ment of science the growth of which is a feature 

 , of the period surveyed on this occasion. 

 ! Stocktaking at the present moment is, however, 



