October 27, 1898] 



NATURE 



6'M 



of a I 



ignorant though we be . . . law," and with the addition 

 reference to his own experiments, i.e. to those on cross- 

 fertilisation. 



This is the most strongly worded form of the Law, and 

 one which is generally adopted. But shortly after the public- 

 ation of the " Origin," i.e. in 1862, the Law took a vaguer 

 form in the "Fertilisation of Orchids" (ed. i., 1862, p. 359), 

 where he wrote : '* Nature thus tells us, in the most emphatic 

 manner, that she abhors perpetual self- fertilisation." This form 

 of the Law is adopted in the "Effects of Cross- and Self- 

 fertilisation " (p. 8), where he writes : "If the word perpetual 

 had been omitted, the aphorism would have been false. As it 

 stands, I believe that it is true, though perhaps rather too 

 strongly expressed." 



The aphorism is clearly not a literal statement of fact, and 

 in describing it as "true," he probably meant that perpetual 

 self- fertilisation is very strongly and very generally guarded 

 against in nature. For he well knew that "some few plants 

 seem to be invariably self- fertilised " (" Cross- and Self- Fertilis- 

 ation," p. 3). With regard to these cases he makes the just 

 remark : " These exceptions need not make us doubt the truth 

 of the above rule, any more than the existence of some few 

 plants which produce flowers, and yet never set seed, should 

 make us doubt that flowers are adapted for the production of 

 seed and the propagation of the species." 



It is only fair to add that this argument also occurs in the 

 " Variation of Animals and Plants under Domestication " (vol. ii. 

 p. 91, 1868), and was therefore of considerably earlier date than 

 his book on " Cross- and Self- Fertilisation" (1876). 



To sum up : 



(i) If the expression, Knight-Darwin Law, is to continue in 

 use, it ought to be applied to a statement on which Knight and 

 Darwin are undoubtedly agreed, i.e. that "nature intended 

 that a sexual intercourse should take place between neighbouring 

 plants of the same species." 



But the name of Knight-Darwin Law is now firmly associated 

 with the positive statement " that no organic being fertilises 

 itself for an eternity of generations," and it would be useless to 

 suggest a new nomenclature. 



(2) If we are to take a Darwinian version of the Law, it seems 

 to me fairer to take the form, "nature abhors perpetual self- 

 fertilisation," which my father adhered to in his later books. 



An example of what seems to me the misuse of the Knight- 

 Darwin Law occurs in my friend Mr. Willis' excellent book, 

 "Flowering Plants and Ferns" (vol. i. p. 46). "In Myrme- 

 codia, &c., Burck has found crossing absolutely prevented, the 

 flowers never opening. Hence the Knight-Darwin hypothesis 

 must be abandoned." If the abandonment of the hypothesis 

 means the recognition of cases of apparently continuous self- 

 fertilisation, the abandonment was made in 1868 by Darwin 

 himself, as I have already shown. But Willis' abandonment 

 seems to me part of an implied contention that Charles Darwin's 

 generalisations are no longer a sufficient basis for floral 

 biology. He seems to think that if the Knight-Darwin Law is 

 not true, the fundamental principles underlying the study of the 

 mechanism of flowers must be sought elsewhere than in Charles 

 Darwin's works. In this point of view I think he is mistaken. 



The attitude of the earlier writers towards the problem of 

 cross-fertilisation seems, if I may venture to say so, to be else- 

 where rather hastily treated by Willis (/oc. cit., p. 45). Take 

 the following passage : ' ' The advantages of cross-fertilisation 

 are often great, and frequently enormous, and as at the first 

 glance they appear to be obtained at little or no cost, we are 

 inclined to expect this method of propagation to prove almost 

 universal. The earlier workers at this subject in fact set out 

 with the idea that cross-fertilisation was, so to speak, the 

 primary object of a flower's existence, whilst self-fertilisation 

 was actually harmful." Almost the whole of this seems to me 

 to be unintentionally misleading. 



That all "earlier workers " did not consider cross- fertilisation 

 the primary object of a flower's existence, is shown by the 

 following passage from "Cross- and Self-Fertilisation," p. 3: 

 "We should always keep in mind the obvious fact that the 

 production of seed is the chief end of the act of fertilisation ; 

 and that this end can be gained by hermaphrodite plants with 

 incomparably greater certainty by self-fertilisation than by the 

 union of the sexual elements belonging to two distinct flowers. 

 Again, reviewing in 1876 (" Cross- and Self-Fertilisation," p. 8) 

 his own treatment of the question in the " Fertilisation of 

 Orchids" (1862), Darwin says: " I should have added the self- 



evident proposition that the propagation of the species, whether 

 by self-fertilisation or by cross-fertilisation ... is of paramount 

 importance." Willis, therefore, seems to me completely wrong 

 if he includes Charles Darwin among the earlier who considered 

 cross-fertilisation the primary object of a flower's existence. 



Nor, I think, is self-fertilisation ever treated by Darwin as 

 positively harmful, though /er/e/wa/ self- fertilisation is so treated. 

 Self-fertilisation is constantly and correctly considered as less 

 advantageous than cross-fertilisation — and in this sense (always 

 bearing in mind the paramount importance of fertilisation of 

 some sort) it may be said that self-fertilisation is relatively 

 harmful. 



Whatever may be the case with other naturalists, Darwin was 

 certainly not inclined to expect cross-fertilisation to prove almost 

 universal. Speaking of orchids, he says ("Fertilisation of 

 Orchids," ed. i. p. 359) : " Considering that the anther always 

 stands close behind or above the stigma, self-fertilisation would 

 have been an incomparably safer process than the transportal of 

 the pollen from flower to flower. It is an astonishing fact that 

 self-fertilisation should not have been an habitual occurrence." 

 He saw clearly that plants pay a price for being so constructed 

 that cross-fertilisation is possible ; in fact, he saw that the 

 evolution of the flower is the result of a gain and loss account 

 between the advantage of cross-fertilisation and the risks and 

 injuries consequent on the flower being open instead of closed, 

 and therefore chasmogamic instead of cleistogamic. And 

 this is in all essentials the theory which Willis ("Flowering 

 Plants and Ferns," p. 46) gives as Macleod's, and proposes as 

 a basis for floral biology, when the Knight-Darwin Law has 

 been abandoned, and H. Miiller's theory also given up. I am 

 not able to read Macleod in the original Dutch, but it 

 would appear from Willis's paper in Science Progress, 

 1895, that Macleod's contribution to the subject is full of valu- 

 able matter, but the essence of his theory (as given by Willis) 

 seems to me to contain nothing with which my father wa^ not 

 familiar. What I object to is the tendency to condense Charles 

 Darwin's contribution towards floral biology to a Knight- 

 Darwin Law, and then, when the abbreviated statement does 

 not explain everything, to abandon — not so much the law — but 

 the general point of view which can only be gathered from 

 Darwin's books as a whole. 



The fact is that some modern biologist uses the Knight- 

 Darwin Law in an inverted way, i.e. in a manner the reverse 

 of Charles Darwin's way of using it. It was not to him a basis 

 for the investigation of floral structures, but a generalisation 

 extracted from that subject to serve as a foundation for the study 

 of wider questions, such as the origin of sexuality. This is clearly 



passagi 



" ...u„ 



Te from the first edition of the " Fertilisation 



NO. 



:)'J. 



VOL. 



58] 



shown in a 



of Orchids," where, after enunciating nature's abhorrence of 

 perpetual self-fertilisation, Darwin goes on ("Fertilisation of 

 Orchids," 1862, p. 359): "This conclusion seems to be of 

 high importance, and perhaps justifies the lengthy details given 

 in this volume. For may we not further infer as probable . . . 

 that some unknown great good is derived from the union of 

 individuals which have been kept distinct for many generations." 

 H. Mliller, perhaps, understood my father's use of the Law 

 when he said (" Fertilisation of Flowers," p. 22) that the 

 Knight-Darwin Law is not necessary for the elucidation "of 

 the forms of flowers." But he would hardly have said as much 

 of Knight's statement, that hermaphrodite flowers are adapted 

 for intercrossing — which is the very foundation of the science of 

 floral mechanism. 



I now pass on to another writer— Knuth — who, in his useful 

 " Bliitenbiologie," seems also to be open to criticism in his 

 treatment of the Knight-Darwin Law. In speaking of H. 

 Miiller's great work, he says ("Bliitenbiologie," vol. i. 

 p. 25): "The laws of Knight, Darwin, Hildebrand, Delpino 

 gave no explanation of the numerous cases of efficacious self- 

 lertilisation, nor of cleistogamy." Here Knuth does not seem 

 to remember the conditions of thought under which the Knight- 

 Darwin Law came into existence. As Loew (" Einfiihrung in 

 die Bliitenbiologie," p. 143) has well said, self- fertilisation was 

 formerly assumed to be the rule in hermaphrodite plants. In calling 

 attention to the existence and importance of cross-fertilisation 

 in hermaphrodites, Knight and Darwin assumed the existence of 

 self-fertilisation. From the point of view of floral biology the 

 important thing was the recognition of cross-fertilisation, and 

 the law in which, unfortunately, this conclusion has been en- 

 tangled need not " explain " the facts which the framers of the 

 law assumed to be a part of common knowledge. With regard 



