January 23, 1896] 



NATURE 



271 



The condition described by Mr. Hill in Perameles is one of 

 natural advance upon the Phascolarctus type, and the facts in 

 proof of the intimacy of relationship between the Placentalia and 

 Lower Mammalia now overlap, like those bearing on birds and 

 reptiles — i.e. just as ArcJucopteryx may be regarded as an avian 

 reptile, and the Odontorinthes as reptilian birds— so the Phascol- 

 arctus may be regarded as a Placental Marsupial, and those 

 placentalia which develop a provisionally vascular ryolk-sac 

 extending to the serous-membrane as Marsupial if not 

 Monotrematous Placentalia. Mr. Hill's discovery, coming close 

 upon that of Woodward that the young of the Diprotodontia 

 are found to develop at fewest five pairs of upper incisors, and of 

 Thoma.s that a representative of the Epanorthida? survives in 

 South America, and at a time when the tooth-genesis of both 

 Marsupialia and Placentalia is receiving exceptional attention, 

 is as welcome as important, in assisting us to form a clearer con- 

 ception of the inter-relationships between these mammals. 



Mr. Hill's observation would appear to lend not a little support 

 to the conclusion which has for years been steadily gaining ground 

 (cf. Nature, vol. xl. p. 420), that the allantoic placenta was 

 primitively discoidal. In an accompanying letter, he informs 

 me that he has more recently come into possession of a uterus 

 containing an unattached blastodermic vesicle ; and it is sincerely 

 to be hoped that he will be able to furnish observations bearing 

 directly upon the important question of the supposed primitively 

 chorionic nature of the Mammalian placenta. 



G. B. Howes. 



Royal College of Science, London, January 14. 



The Origin of Plant Structures. 



May I call attention to a serious omission, and reply to one 

 or two points, in Mr. Barber's review of my book ? He says : 

 •'Of the inheritance of such acquired characters there is no 

 proof at all. We are offered instead the ' argument of coin- 

 cidences ' and the ' cumulative evidence of probaVjilities, which 

 amounts to a moral conviction.' Clearly, before rejecting a 

 well-established and widely applicable hypothesis, something 

 more tangible is required " (Nature, December 19, 1895, 

 p. 145). 



First, with regard to inheritance. Mr. Wallace also asked 

 for some proof of this ; and I reply again that nature herself 

 supplies it ; for plant structures are reproduced by seed every 

 year. It is the previous question, "How have they arisen?" 

 with which I am concerned. But I have, in fact, given plenty 

 of cases : as in mv experiment with Ononis, Buckman's 

 parsnip, P'lahoull's with alpine plants, &c. 



Moreover, the objector should state whether he means that 

 any altered features in a plant should be reproduced by seed 

 irrespective of the environment, or not. If a plant changes 

 under new conditions — as ample experience shows it may — of 

 course all its offspring will follow suit, under the same environ- 

 ment, irrespective of heredity; and if the conditions be maintained 

 long enough, then the new features will tend to become relatively 

 fixed, as all cultivators know. As long, however, as a natural 

 environment is constant, no varieties are, as a rule, to be 

 expected. Under cultivation, this rule does not hold good. 

 Thus, e.g. Brassica oleracea gives rise to no varieties in nature ; 

 but there are very many fixed and hereditary races in artificial 

 soils. 



Secondly, the truth of an hypothesis or deduction cannot be 

 more surely established than by " verification by experiment.'* 

 Thus, with desert, aquatic, alpine, maritime, and other plants, I 

 not only established the truth of my contention by induction, 

 but have given the experimental verifications both of others and 

 myself. For //»V fact Mr. Barber gives me no credit. It is in 

 these two lines of proof, viz. by iiiduction and experiment, that 

 the theory of " the origin of species by means of natural 

 selection '" is wanting. It is, as Mr. Barber says, based on an 

 " assumption," and is an a priori deduction that, because plants 

 can vary indefinitely under cultivation, therefore they do so also 

 in nature. This has never been verified. 



Mr. Barber adds : " It is usually agreed that, from the nature 

 pf the case, a definite proof of the action of natural selection 

 is difficult, if not impossible, in the present state of our 

 knowledge." Is not this a most damaging admission? If 

 the word " present" has to cover the thirty-six years since the 

 " Origin " appeared, it would seem to be about time to abandon 

 the theory even as a working hypothesis. The " wideness of 

 its application " is no test of the truth of a deduction ; for 



NO. 1369, VOL. 53] 



though natural selection may account for all organic structures, 

 it can only do so because it is assumed that it can account for 

 them. Herein it agrees precisely with the theory of special 

 creations ; which is equally assumed to be capable of accounting 

 for every organic structure. 



Mr. Barber tells me that I show a want of good taste in 

 "narrowing" Darwin's field of observation. I much regret 

 that anything I have written should be regarded as uncourteous ;. 

 but it is Darwin himself who admits the " imputation," 

 for he wrote: "I will give in detail all the facts which I have 

 been able to collect," «.^. of "definite action" in plants, and 

 mentions about thirty instances which he had heard of. He did 

 not believe in definite variation being the rule in nature. 



Lastly, I make no claim " to reconstruct the theory of evolu- 

 tion." All I have done is to take the following passage of 

 Mr. Herbert Spencer's Essay on " The Development Hypo- 

 thesis" (published in 1852, seven years before Darwin's 

 " Origin " appeared) as my subject ; and I have simply verified 

 its profound truth in its application to plants. " The supporters 

 of the developmental hypothesis can show . . . that any existing 

 species — animal or vegetable — when placed under conditions 

 different from its previous ones, immediately begins to undergo 

 certain changes of structure fitting it for the new conditions . . » 

 that in successive generations these changes continue until ulti- 

 mately the new conditions become the natural ones. . . . They 

 can show that throughout all organic nature there is at work a 

 modifying influence of the kind they assign, as the causes of . . . 

 specific differences ; an influence which, though slow in its- 

 action, does in time, if the circumstances demand it, produce 

 marked changes." George Henslow. 



I AM sorry I cannot agree with Prof. Henslow as to the 

 nature of the proof required of him. 



The fact that " plant structures are reproduced by seed every 

 year" in nature is surely no proof of the inheritance of acquired 

 characters ! 



Plants are, it is true, exceedingly plastic structures, and, as 

 all allow, are both temporarily and permanently affected by 

 their surroundings. This, if we needed further proof, Prof. 

 Henslow has repeatedly demonstrated by his interesting series 

 of facts. There seems to be danger, however, of confusing 

 these changes in the individual with changes in the race. Prof. 

 Henslow makes the former the prelude to the latter ; and the 

 first question to be settled is, What connection is there between 

 these two classes of changes? In other words, Are acquired 

 characters hereditary ? 



The issue is the same if we seek for the causes of variation. 

 Darwinism, realising that there is a gradual adaptation of plants 

 to altered surroundings, explains the fact by the indirect 

 influence of the environment acting through natural selection. 

 Among plants, which are stimulated to vary in all directions 

 under change of conditions, those are preserved which vary so 

 as to place themselves in adaptation to their new surroundings. 

 Prof. Henslow substitutes the direct influence of the environ- 

 ment upon the individual plant, and asserts that the changes 

 thus induced "become relatively fixed if the conditions are 

 maintained long enough." Here is an assumption that the ciaw^^? 

 in the race is the outcome of the direct effect of the environment 

 upon the individual, or, again, that acquired characters are 

 hereditary. I have carefully re-examined the cases mentioned 

 of Ononis, parsnips and alpine plants, but cannot trace any proof 

 of this assumption. 



As to the indefinite variation of plants and animals in nature, 

 it is difllicult to conceive of doubt upon the subject. It is a 

 common saying that " no two blades of grass are alike," although 

 conditions could hardly be imagined more uniform than those 

 in one and the same field. Moreover, the fact of indefinite 

 variation has been fully proved by Prof. Wallace in his work on 

 Darwinism. In chapter iii. , on " Variability of Species in a State 

 of Nature," the whole subject has been exhaustively dealt with, 

 and I cannot do better than refer Prof. Henslow to that chapter, 

 where numerous cases are given, both in animals and plants. I 

 would especially refer to the extracts from Darwin's note-books 

 there published for the first time. 



I do not quite follow Prof. Henslow in rejecting evidence 

 drawn from cultivated plants. Placing wild forms under cultiva- 

 tion is a severe change of environment, and any such change 

 induces, of itself, a great tendency to vary. "The influences 

 which work slowly in nature are intensified ; and the substitution 



