July 7, 1910] 



NATURE 



an organism, how is it to be adapted to a different 

 environment? The adjustment may be effected by 

 further complication or by simplification. It may be 

 noted that in regard to the latter there is a close 

 parallel in the evolution of machinery. Whole trains 

 of mechanism are continually being swept away with 

 an increase of efi'iciency. Compare, for example, a 

 turbine with a marine engine. Here structural retro- 

 gression has made for functional progression. We 

 owe it to Lankester for pointing out that "degenera- 

 tion " is really simplification leading to closer adapta- 

 tion. Progress in biology is not ethical, but position 

 in the phyletic scale. The last of the Plantagenets is 

 said to have kept a turnpike ; but he may have been 

 not the less authentic. 



The instability of prize-bred domesticated races re- 

 quires careful scrutiny. The late Duke of Devonshire 

 pointed out to Lankester that racehorses are bred for 

 speed and not for "points." The conclusion that I 

 draw from Sir Walter Gilbey's facts is that breeders 

 have not yet succeeded in fixing this particular quality. 

 But short-horns, which are bred for points, have 

 reached a high degree of stability ; if they had not no 

 one would give a thousand guineas for a bull. The 

 purchase of a possible racehorse is confessedly a 

 gamble. For my own part, I am content with Lan- 

 kester 's view that nature " with remorseless thorough- 

 ness " can throw overboard hereditary tendencies, if 

 it is advantageous to do so ; and this is really the 

 same thing as Dr. Archdall Reid's selection of retro- 

 gfressive germ-plasms, except that he throws on 

 natural selection the burden of defeating its own aim. 



Apart from speculation, we have in Gallon's law of 

 regression to mediocrity an empirical result which is 

 perfectly general inasmuch as it deals impartially with 

 excess and defect. It produces "a sensible stability of 

 type and variation from generation to generation." 

 It has always appeared to me the most important 

 positive addition to the Darwinian theory, and it has 

 seemed possible that it would open the door to a 

 mechanical explanation of retrogression, or, as I 

 prefer to say, of simplifi,cation ; and this is apparently 

 in .Archdall Reid's mind, as he remarks that "regres- 

 sion is but the first phase of retrogression," though 

 he has not followed it out further. Regression is 

 independent, apparently, of natural selection, while 

 retrogression is not. 



This leads to another point which is often over- 

 loolted. The mere " maintenance of a structure " is 

 dependent on the continued action of natural selec- 

 tion. As Poulton insists, it is by its operation that 

 "all functional parts of an organism are kept up to a 

 high standard." It may be a private heresy of my 

 own, but I can attach no more meaning to the 

 "cessation" and "reversal" of selection than if those 

 terms were applied to gravitation. 



The chapter on Mendel's laws is altogether admir- 

 able. It is probably the most luminous account of 

 them which has been published. " There can be no 

 doubt of the actual occurrence of the Mendelian 

 phenomena. We must, endeavour, therefore, to esti- 

 mate the part played by them in nature." Now where 

 species or stable varieties are crossed we get 

 simple blending, as in the Mulatto. "Mendelian 

 NO. 2123, VOL. 84] 



reproduction is one of the rarest things in nature." 

 " Mendelian traits . . . are common only when arti- 

 ficial varieties . . . are crossed by man." It would be 

 impossible with any justice to attempt to summarise 

 the argument. The majorit)' of Mendelian traits "are 

 concerned with reproduction." The illuminating con- 

 clusion, in which, however, the author finds himself 

 anticipated by T. H. Morgan, is reached that they are 

 analogous to sexual characters which are alternative, 

 i.e. are latent or patent in the opposite sex. If this 

 explanation holds good, and it has the obvious merit 

 of including phenomena not obviously connected at 

 first sight, it effectually disposes of "segregation"; 

 and "unit-characters" necessarily follow. Hut their 

 existence had already become precarious, for Prof. 

 Karl Pearson kindly informs me that he has entirely 

 failed to discover any which, to put it briefly, can be 

 described as having unitary properties. It is pointed 

 out that the inheritance of mutations is alternative, 

 and the inference is drawn that characters which blend 

 in crossing cannot have arisen as mutations. 



Lastly, we come to the " Function of Sex." This is 

 found to be an adaptation " to blend parental char- 

 acters." Further, it is concluded that "blending, with 

 its swamping effects . . . eliminates useless char- 

 acters and variations." This at once explains retro- 

 gression, and at bottom on this head there is probably 

 not much difference between Lankester and the 

 author. Mutations are alternative and Mendelian ; 

 fluctuations are blended; whence Galton's law of re- 

 gression and stability at once follows. "The average 

 experience of the whole race . . . becomes the deter- 

 mining factor in evolution." 



Two incidental points deserve notice. Partheno- 

 genesis " occurs as a rule amongst simple forms." 

 But it is found to occur much more frequently than 

 was supposed amongst flowering plants ; the dandelion 

 is an example. Still, it may be presumed that sexual 

 reproduction and cross-fertilisation occasionally occur. 

 Fertility, both on biometric and general grounds, is 

 thought to be a transmissible adaptation. Karl Pear- 

 son has, however, arrived at the important conclusion 

 that there is "little or no demonstrable inheritance of 

 fertility." Further, he is " forced to the conclusion 

 that the smallness of the hereditary factor in fertility is 

 an essential feature of Darwinian evolution." It is 

 interesting to note that in this case deductive reason- 

 ing has led to diametrically opposite conclusions. 



This disposes of the first part of the book. I do 

 not know that I have come across anything more sug- 

 gestive on the subject since the "Origin" itself. It 

 may be added that Prof. Turner has thrown the main 

 argument into a quasi-mathematical shape in the 

 appendix. The latter and larger portion of the book 

 is difficult to review in any reasonable space. It is a 

 striking commentary on the contention of de Vries 

 that organic evolution has nothing to say on social 

 problems. It ranges over a wide field, including even 

 a short svstem of philosophy, and will probably be 

 found the more interesting because the least technical, 

 and might well have been published separately. 



Disease and immunity are admirably discussed. 

 Races become tolerant through selection working on 

 germinal variation. Protoplasm learns to neutralise 



