Feb. 2 1, 1889] 



NA TURE 



389 



the germ-cells in this altered individual, which were originally 

 continuous with the germ-cell from which the individual itself 

 developed, remain entirely unaffected by the action of the 

 conditions on the body, and when they begin to develop have 

 exactly the same properties as the germ-cell in the generation 

 preceding. 



Heredity, then, according to Weismann, is simply the pro- 

 perty possessed by a germ to develop into exactly the same 

 type in each successive generation. He says : " Ich stelle mir 

 vor, dass die Vererbung darauf beruht, dass von der wirksamen 

 Substanz des Keimes, dem Keimplasma, stets ein Minimum 

 unveriindert bleibt," &c. 



As the action of conditions can give rise to no hereditary 

 individual differences, these must be due to some other cause. 

 This cause, Weismann says, is to be found in sexual, or, as 

 Haeckel calls it, amphigonous, reproduction. Sexual reproduc- 

 tion consists in the fusion of two complementary germ-cells or 

 of their nuclei : each of these germ-cells has a specific mole- 

 cular structure, on which depend the hereditary tendencies of 

 the organism whence the germ-cell is derived. Thus, in fer- 

 tili^tion, two hereditary tendencies are mingled, and thus the 

 offspring does not resemble exactly either of its parents, but 

 combines the characters of each together. 



In order that there shall be no ambiguity about his argu- 

 ment, Weismann precisely states what, according to his view, 

 and, as he believes, in actual fact, occurs in monogonous repro- 

 duction, i.e. in parthenogenesis, where there is only one parent 

 instead of two. If, in a species reproducing parthenogenetically, 

 all the individuals were perfectly similar, all the descendants 

 throughout any number of generations would continue similar, 

 leaving aside evanescent differences due to conditions, and which 

 are not hereditary. In such a case no selection, Weismann 

 says, would be possible, and therefore no evolution in any 

 direction. 



" Processes of selection in the proper sense of the word, 

 those which produce new characters by the gradual increase of 

 characters already present, are not possible in species which 

 reproduce asexually." 



"If it were once proved that a species reproducing itself 

 solely by parthenogenesis had been transformed into a new 

 species, thereby it would be proved at the same time that other 

 causes of modification exist than processes of selection, for by 

 selection the new species could not have been formed." 



But with sexual reproduction it is quite otherwise. Weismann 

 points out that no two individuals of different generations could 

 ever be similar where reproduction is sexual, and even the in- 

 dividuals of one family, born of the same two parents, would 

 not be similar, because the various tendencies in the parents are 

 present in different intensities at different times, though he gives 

 no reason for this assumption. 



According to Weismann, the individual hereditary differences 

 so produced are the basis on which selection acts ; and these 

 differences thus explained, summed up or combined in different 

 ways by selection, give a complete and satisfactory explanation 

 of all organic evolution. 



Now, let us examine this theory a little. P"or the sake of 

 simplicity we will in most cases consider the effect of the sup- 

 posed processes on one organ. In the first place, what ground 

 is there for assuming that Vermischung would ever cause an 

 increased development in the offspring of an organ possessed by 

 the parents ? Heredity, as understood by Weismann, is nothing 

 more than the property in the germ-cell of developing into an 

 individual like that from which it was derived. If each parent 

 possessed a given organ in the same degree of development, a 

 degree unaffected by external conditions, then both the ovum 

 and the spermatozoon will, on this view of heredity, have the 

 property of developing into an individual with the same organ 

 developed to the same degree. When the two properties are 

 combined by fertilization, the fertilized ovum ought to have the 

 property of developing this character with still greater certainty, 

 but why should it have the property of developing the character 

 to a higher degree than that reached in either parent? By 

 Vermischung , in its literal sense, the union of the two here- 

 ditary powers cannot have this effect. If by Vermischung 

 it is meant that the offspring is intermediate between the 

 two parents, then the mean of two equal characters is the same 

 character again. And if this is what Weismann means by 

 Vermischung, then a character developed to a certain degree 

 in one parent, and not at all in the other, would in the offspring 

 be developed to exactly half the degree in which it existed in 



the one. And so on. But it is obvious that in this way no 

 increase of any character could ever occur. 



But of course Vermischung may mean something else. It 

 may mean that the hereditary powers of ovum and spermatozoon 

 are added together, that the result of copulation between the 

 germ- cells is not the mean, but the sum, of the properties of both. 

 In this case, evolution would be extremely rapid, for each child 

 would be equal to both its parents rolled into one. If each 

 parent, say, among cattle, had horns equally well developed, the 

 offspring would have horns twice as big. And it is obvious that 

 in this way no decrease could ever occur, for if one parent had 

 an organ developed and the other had no trace of it, the off- 

 spring would have it in the same degree as the one. 



Now, it seems to me that, if Vermischung does not mean 

 either of these things, there is only one other meaning it can 

 have, and that is, that the hereditary powers of the copulating 

 germ- cells reinforce one another to some extent, but not to such 

 an extent that the result is equal to their sum. If this be the 

 meaning, then there can never be any decrease in a character 

 once formed. For, if every individual of a species possesses a 

 certain organ, let us take the hind-legs in a mammal, then if two 

 individuals which have these organs less developed than any 

 other individuals in the species, copulate, the offspring resulting 

 must have hind-legs better developed than either of them. Thus 

 the whale could never have been evolved. 



It follows, therefore, that, on Weismann's theory of variation, 

 evolution is impossible. And as acquired characters are not 

 inherited, no other theory of variation can be discovered. 

 Therefore evolution is impossible altogether : the extremes meet, 

 and the Darwinian principle overstrained goes rather to prove 

 the fixity of species than their plasticity. 



J. T. Cunningham. 



Mr. Howorth on the Variation of Colour in Birds. 



Allow me to assure Mr. Howorth that I have no theory to 

 maintain. I simply called attention (supra, p. 318) to an over- 

 looked hypothesis, propounded long ago, and, so far as I know, 

 still unrefuted. Neither have I any wish to argue the question. 

 Indeed, controversy about it is happily almost impossible, since he 

 admits the chief fact of which I reminded him to be what he 

 now terms {supra, p. 365) "an elementary postulate "—an ex- 

 pression far stronger than I should venture to use ; but had he 

 before shown any disposition to recognize it, my remarks had 

 not been written. On the contrary, he implied {supra, p. 294) 

 that it was a recent discovery, as it certainly appears to have 

 been to him. I trust he will excuse me for having pointed out 

 its want of novelty, just as he seems to excuse Prof Geikie for 

 pointing out the antiquity of his views as to the former climate 

 of Siberia ; and at the same time I have to ask Mr. Howorth's 

 pardon for demurring to some of the assertions in his last com- 

 munication, especially that as to the avifauna of Siberia having 

 been " worked out from end to end." I dare not hope to see 

 the day when this shall be done ; but then I am not of a 

 sanguine temperament. 



I take this occasion to mention thit in line 3 of the second 

 paragraph of my former letter (p. 318) the word "Russian" 

 was omitted before " explorers and naturalists." Of course it 

 will be understood to cover Poles, as well as all those foreigners 

 who were employed by the Russian Government. 



Alfred Newton. 



Magdalene College, Cambridge, February 16. 



Currents and Coral Reefs. 



May I be allowed space to call attention to a remarkable fact 

 relating to the growth of coral reefs, which has apparently (as 

 far as I can ascertain) had no explanation, and which might 

 assist materially in the elucidation of some problems relating to 

 ocean currents about which— although the broad facts are known 

 — a great deal of doubt exists ? It might also give us some 

 idea of the flow of submarine currents, the direction of which it 

 is very difficult to determine. 



It will be observed that in all coral formations there are in 

 some places remarkable extensions of them from the land, which 

 is not accountable for by supposing the depths only to be 

 shallower in those directions, and the only alternative we have 

 then is that the food supply must come chiefly from that direc- 

 tion, and this supply could only be kept up by currents striking 

 the reef at these points. To give an instance of this, I might 



