NATURE 



12> 



THURSDAY, MAY 23, 1907. 



MENDELISM. 

 Mcndclisni. By R. C. Punnett. Pp. vii + 84. 

 (Cambridge : Macmillan and Bowes ; London : 

 Macmillan and Co., Ltd., 1907.) Price 2s. net. 



A SECOND edition of Mr. Punnett's" Mendelism " 

 lias followed tlie first after an interval of two 

 years. The book gives a very good account of 

 Mendelian worl<. Lssuing as it does from Cam- 

 bridge, the source of by far the greater part of the 

 Mendelian discoveries of the last six years, it is the 

 most authoritative account of this subject, and as it 

 is eminently readable it is the very book for anyone 

 who wants to know what Mendelism is. It is cheap 

 and of a very convenient size, and we cannot suppress 

 an expression of our admiration for the beautiful 

 purple colour of the cover of the second edition. 



At the same time, no good can be done by refusing 

 to face the fact that the truth of the Mendelian 

 doctrine is not universally admitted. No one denies 

 the extraordinary interest of these discoveries. He 

 must be a very callous man who is not fascinated by 

 the way in which the proportions 9:3:4 and 9 : 7 

 and the phenomenon of reversion in certain cases 

 have all been brought into line. But we hold that 

 he must be a very rash man who accepts without 

 further question the doctrine of gametic purity. Yet 

 it is just in the sphere of interpretation that 

 Mendelians are so certain. Once in this sphere, we 

 can no longer be guided by facts — if we were dealing 

 with facts we should be in the sphere of discovery — 

 but by "such things as our mind conceives." And 

 one's attitude should be one of continual, unceasing, 

 and active distrust of oneself. The attitude of the 

 Mendelian is different from this. He may reply that 

 he is only triumphant about his discoveries; but we 

 must remember that there is no fixed criterion by 

 which we can say where discovery ends and inter- 

 pretation begins ; and we must be careful not to 

 hesr the question by defining discoverv as that about 

 which there can be no doubt. 



There are those who deny the theory that the germ 

 cells of an extracted recessive are pure in respect of 

 the character of the organism which contains them, 

 and who assert that the characters of the hybrid 

 which produced it are not absent from, but latent in, 

 those germ cells. If this were discovered to be the 

 case, it would be regarded as a demonstration of the 

 falsity of the doctrine of gametic purity by everyone 

 who was not a Mendelian. But we should strongly 

 condemn the proclamation of such a conclusion, be- 

 cause we think it is high time that the spirit which 

 derives satisfaction from the victory of one opinion 

 over another should be swept from science. There 

 is no place for the party system in science ; because 

 it tends to make the triumph of truth the main object 

 and truth itself a secondary one. We are not arguing 

 that Mendelian theory is untrue, but that the attitude 

 of anyone daring to say of anything " this is true " 

 should be apologetic rather than victorious. 



There is another and a larger point of view from 

 XO. IC,6o, VOL. 76] 



which we may examine the Mendelian position : it is 

 that which refers to the relation between the 

 Mendelian and the material with which he deals. 

 The differences between biomctrician and Mendelian 

 have been due partly to the fact that these two sets 

 of workers have dealt with different sets of facts. 

 But we are concerned with the difference between 

 their attitudes to the same classes of facts, and with 

 the paradox that in spite of this difference they both 

 claim to have introduced exact methods into biology. 

 How is it that the two schools which claim to have 

 introduced the exact method into the study of biology 

 are not at peace? What is the relation between the 

 methods of the two schools? One author attempts 

 to express the difference in the statement that the 

 Mendelian deals with units and the biometrician with 

 masses, and illustrates this view by saying that the 

 difference between the relation of the biometrician 

 and that of the Mendelian to the units with which 

 they deal is the same as the difference between the 

 relation of the physicist and that of the Maxwellian 

 demon to the units (the atoms) with which they deal. 

 The physicist and the biometrician deal with them in 

 masses. The demon and the Mendelian deal with 

 them separately. It is striking testimony to the 

 callousness of biologists to general discussion that it 

 has never been pointed out that this comparison, 

 though plausible, is based on a fallacy. To anyone 

 who tries to take a broad view of the matter, the 

 truth or falsity of the statement (repeatedly made by 

 Mendelians) that the biometrician deals with masses 

 while the Mendelian deals with units is the most 

 interesting question presented by this whole subject. 

 For if the Mendelian really does deal with units 

 while the biometrician deals with masses composed 

 of these units, and if the Mendelian sets out with 

 the object of enabling himself to predict what will 

 be the result of a given union, and succeeds, while 

 the biometrician starts on the assumption that a 

 knowledge of the ancestry of a given pair does not 

 enable him to predict the character of its offspring, 

 there is little to be said for the " application of exac* 

 statistical methods to the problems of biology." 



But is it really true that the Mendelian deals with 

 the units of which the biometrician 's masses are 

 composed? We believe not. In order to see what 

 the real state of affairs is we must try to begin at 

 the beginning. The difference between the two- 

 schools lies in the difference between their respective 

 attitudes to natural phenomena. The biometrician 

 says, " We look at them as close as we can and 

 we see nothing uniform." The Mendelian, " We 

 look as close as we choose and we see everything 

 uniform." The latter does not pretend that " dwarf " 

 peas are not variable, but treats them as if they w-ere 

 all the same. The former does not pretend they are 

 not all "dwarf," but treats them as if they were all' 

 different. 



The exactness of the biometrician makes him 

 count the number of hairs per square centimetre on- 

 the lower leaf surface of Lychnis vespertina ; the 

 exactness of the Mendelian enables him to tell at a 

 glance in a row of hybrid stocks which are hoary 



E 



