January 9, 1908J 



NATURE 



237 



have been no antecedent separation of factors. Therefore 

 the evidence is that mutations tend to be inherited in the 

 mode of sexual characters. There are, however, differ- 

 ences. Se.xual characters tend to alternate more perfectly, 

 to cohere together in their respective sets more closely, 

 and to have a lesser tendency to blend with their opposite 

 numbers than Mendelian characters ; but this is only what 

 might be expected, for the mode of inheritance for sexual 

 characters has been established by stringent selection. 

 Some sexual characters, however, sometimes blend or 

 (.-hange places with their opposite numbers, as in so-called 

 human hermaphrodites ; the dominance of some Mendelian 

 characters is very imperfect ; reciprocal IVIendelian crosses 

 sometimes produce unlike results ; and the inheritance of 

 some Mendelian characters {e.g. colour-blindness) is sexual. 

 Were their occurrence the rule, not the exception, we 

 would speak of them as sexual characters. Properly 

 speaking, Mendelian characters are non-sexual traits which 

 .are reproduced in the sexual mode. If, however, we 

 examine any list of so-called Mendelian characters, we 

 find that the majority may fairly be described as secondary 

 sexual characters, though not necessarily as sexual differ- 

 ences, for example, colour and form of plumage. It seems 

 clear, then, that there is no real segregation, no real 

 alternative inheritance, but only alternative reproduction, 

 alternative patency and latency. Therefore blending is 

 universal. Unless the reappearance of ancestral traits can 

 be explained, the bottom falls out of the Mendelian hypo- 

 thesis. Nevertheless, Mendelian facts are very valuable, 

 inasmuch as they indicate the difference between natural 

 and artificial selection. The experimental observer is able 

 to note only large diflferences between mating individuals. 

 In practice, he has almost limited his materials for study 

 to domesticated varieties. He has altogether ignored 

 fluctuations. Hence the mutation and Mendelian theories. 



It is possible that mutations are more common amongst 

 artificial than amongst natural varieties, in which the 

 range of variability is more stringently limited. But 

 amongst the entirely natural varieties of the species we 

 know most intimately (man) mutations are common 

 enough, and their reproduction tends to be Mendelian. 

 But all are so injurious in the struggle for existence or 

 for mates that when possible they are treated surgically. 

 Never yet has a useful human mutation been recorded. 

 Man has a written history of thousands of years, and 

 human varieties differentiate whenever geographical isola- 

 tion is sufficiently complete and prolonged. Men are fond 

 of noting wonders. But, notwithstanding the immense 

 range of material, never yet has the origin of a human 

 varietv bv mutation been recorded. It is easy to conceive 

 of evolution as resulting from mutations when we limit 

 our materials of thought to the colours and shapes of 

 flowers and leaves of plants which are preserved under 

 human rare. It is not so easy to think of it as founded 

 on mutations when we take into account the exquisitely 

 co-adapted internal parts of a complex animal, amongst 

 which a mutation w^ould have the same effect as one 

 occurring in one of the parts of a watch. 



The extreme instability of fluctuations has been noted 

 and has furnished a main argument to the supporters of 

 the mutation theory. There is abundant reason for 

 believing that in a blend the retrogressive character tends 

 to predominate. Thus racehorses degenerate unless care- 

 fully selected. Suppose a country in which malaria is 

 prevalent and another from which it is absent. In the 

 former, variations favourable against malaria are selected ; 

 like mates with like ; therefore blending causes little or 

 no retrogression, and the mean of the race is raised in 

 each generation. In the latter, though favourable varia- 

 tions occur, unlike individuals mate ; therefore blending 

 causes retrogression, and the race is rid of a useless re- 

 dundancy. Apply this reasoning to all variations and all 

 characters, and the function of conjugation becomes 

 apparent. It is, in effect, a selective agent of retro- 

 gression. Selection rough-hews the type; retrogression 

 chisels out the finer lines. Reproduction is bi-parental in 

 all the higher and more complex forms in which nature's 

 task of closely adjusting the numerous co-adapted parts 

 is most difficult. 



If we accept the theorv of blended inheritance, we are 

 able to assign a useful function to conjugation. But to 



both the Mendelian and the mutationist, sex is a pheno- 

 menon to be explained away. According to the former, 

 conjugation merely jumbles together elements which may 

 be incongruous. According to the latter, conjugation is 

 nothing other than an obstacle to the survival of muta- 

 tions, which have the best chance of surviving when 

 reproduction is parthenogenetic. It has been said by some 

 Mendelians and mutationists that fluctuations are due 

 merelv to temporary effects of nutriment, temperature, and 

 the like ; mutations alone are permanent. How, then, is 

 it possible to explain the fact that when reproduction is 

 parthenogenetic " thousands of forms may be cultivated 

 side by side in the Botanical Gardens, and exhibit slight 

 but undoubted differentiating features, and reproduce them- 

 selves truly by seed" (de Vries)? This does not happen 

 when reproduction is bi-parental. It can hardly be con- 

 tended that mutations are a Ihousand-fold more numerous 

 when reproduction is parthenogenetic than when it is 

 bi-parental. We are driven to the conclusion that the 

 fluctuating nature of fluctuations when reproduction is 

 bi-parental is due to the retrogression caused by blending. 

 The seeming permanency of mutations is due to their 

 mode of reproduction. 'They take longer to retrogress 

 than fluctuations only because they are bigger. All latent 

 characters, since they are not selected, tend to retrogress. 



UNIVERSITY AND EDUCATIONAL 

 INTELLIGENCE. 



Prof. Adolphe Carnot, latterly director, and for many 

 years professor, of chemistry at the Paris School of Mines, 

 has retired with the title of honorary director of the school. 



It is announced, says .Science, that Mr. Emile Berliner, 

 of Washington, the inventor of the gramophone, has given 

 2500;. as endowment of a research fellowship for women 

 who have demonstrated their ability to carry on research 

 work in physics, chemistry, or biology. 



A RESEARCH scholarship or scholarships, founded by Mr. 

 Andrew Carnegie, will be awarded shortly, irrespective of 

 sex or nationalitv, on the recommendation of the council 

 of the Iron and' Steel Institute. Candidates, who must 

 be under thirty-five vears of age, must apply on a special 

 form before the end of February to the secretary of the 

 institute, 28 Victoria Street, London, S.W. The object 

 of this scheme of scholarships is not to facilitate ordinary 

 collegiate studies, but to enable students, who have passed 

 through a college curriculum or have been trained in 

 industrial establishments, to conduct researches in the 

 metallurgy of iron and steel and allied subjects, with the 

 view of aiding its advance or its application to industry. 



The sixth annual meeting of the North of England 

 Education Conference was opened at the University of 

 Sheffield on January 3, under the presidency of Prof. M. E. 

 Sadler. After the 'presidential address. Prof. Hicks read 

 a paper on the function of a modern university, and Dr. 

 R. H. Crowley (Bradford) and Dr. Clement Dukes 

 (Rugby) dealt with medical inspection of school children. 

 Other subjects discussed were holiday and open-air schools, 

 compulsory attendance at evening schools, and the teach- 

 ing of history. On the following day Sir William Clegg 

 presided, and the morning sitting was devoted to con- 

 sideration of the work of training colleges. The_ after- 

 noon topics were : — (i) the treatment of defective children ; 

 (2) house-craft in girls' schools ; (3) artistic perception in 

 children. 



The London County Council Conference of Teachers was 

 held on January 2, 3,' and 4, when more than 1200 visitors 

 signed the attendance book, and as this was the tenth of 

 these annual meetings, it is clear that their utility is 

 appreciated. With a wise liberality, the County Council 

 promises to send a verbatim report of the proceedings to 

 those who attended the conference, of which a noteworthy 

 feature was the frank cordiality with which the Council 

 inspectors and teachers interchanged views. The principal 

 topics discussed were the place of nature-study in the 

 curriculum, the study of botany by girls, the commercial 

 education of boys, manual instruction for young children, 

 practical suggestions for school library management, and 

 recent pedagogic experiments in the study of literature 

 and of open-air geography. 



NO, T993, VOL. 77] 



