July 20, 1893J 



NATURE 



267 



of vitreous reflection and refraction and of reflection at 

 metallic surfaces. 



The book contains a clear account of the theoretical 

 aspects of the above questions, the mathematical treat- 

 ment beins as elementary as is consistent with the nature 

 of the subject. 



LETTERS TO THE EDITOR. 



{ Tht Editor does not hold himself responsible for opinions ex- 

 pressed by his correspondents. Neither can he undertake 

 to return, or to correspond with the writers of, rejected 

 manuscripts intended for this or any other part of Nature 

 No notice is taken of anonymous communications . ] 



The Non-Inheritance of Acquired Characters. 

 I WISH to call the attention and elicit the opinion of natural- 

 ists as to the interpretation of certain facts bearing upon this 

 question. 



In my article in the Fortnightly Review of May last, p. 664, 

 I give what appears to be a new interpretation of facts which 

 have been often quoted, as to the change in the external characters 

 of a Texan species of Saturnia when the larvae were fed upon 

 /uglans regia, its native food-plant being Juglans nigra; and 

 the somewhat analogous facts as to Artemia salina being 

 changed into A. Milhausenii (the former living in brackish, the 

 latter in salt water) when the water became gradually more salt ; 

 the change in this case being progressive, year by year, and 

 proportionate to the change in the saltness of the water. The 

 reverse change was also effected by gradually reducing the 

 salinity of the water inhabited by A. Milhausenii. 



As regards the former case I remarked in my article as 

 follows : — 



"Prof. Lloyd Morgan (in his 'Animal Life and Intelligence,' 

 pp. 163-166) clearly sees that this and other cases do not prove 

 more than a modification of the individual ; but it seems to me 

 to go further than this. For here we have a species the larvK 

 of which for thousands, perhaps millions, of generations have 

 fed upon one species of plant, and the perfect insect has a 

 definite set of characters. But when the larv^ are fed on a 

 distinct but allied species of plant, the resulting perfect insect 

 differs both in colouration and form. We may conclude from 

 this fact that some portion of the characters of the species are 

 depeudent on the native food-plant, /uglans nigra, and that 

 this portion changed under the influence of the new food-plant 

 Yet the influence of the native food-plant had been acting un- 

 interruptedly for unknown ages. Why then had the resulting 

 characters not become fixed and hereditary? The obvious 

 conclusion is, that being a change produced in the body only 

 by the environment, it is not hereditary, no matter for how 

 many generations the agent continues at work ; in Weismann's 

 phraseology it is a somatic variation, not a germ variation " 



I then referred to the marked difference between somatic and 

 germ variations in plants, the former disappearing at once the 

 latter persisting, when cultivated under abnormal conditions • 

 and also to the cases of many closely allied species of animals 

 and of the races of mankind, which preserve their distinctive 

 characteristics when living and breeding under very different 

 conditions. ' 



The above seems to me a perfectly valid and logical argu- 

 ment, and I was interested to see how it would be met by 

 Lamarckians, who have frequently referred to the same facts as 

 bemg obviously in their favour, though without any attempt to 

 show how and why they are in their favour. I was therefore 

 rather surprised to read, in the July issue of the Contemporary 

 Review, a paper by Prof. Marcus Hartog, in which he charac- 

 terises my argument as a very bad kind of special pleading, and 

 adds that It amounts to this: " Any change in the ofi-spring 

 produced by altered conditions in the parent is limited to 

 characters that are ' not fixed and inherited' ; for fixed and in- 

 herited characters cannot be altered by changed conditions in 

 the parent; therefore no experimental proof can be given of 

 the transmission of acquired characters." 



The above is of course simple reasoning in a circle, and I 

 cannot recognise it as my reasoning. I have made no general 

 proposition that ''fixed and inherited characters cannot be 

 altered by changed conditions in the parent," or that "no ex- 

 perimental proof can be given of the transmission of acquired 

 NO. 1238, VOL. 48] 



characters." But I argue that when a decided character is 

 immediately changed by changed conditions of the individual, 

 as in Saturnia, it is not "fixed and inherited." The experi- 

 ment itself shows that it is not a fixed character, and there can 

 be no proof that it is inherited so long as it only appears under 

 the very same changed conditions that produced it in the 

 parent. 



As to experimental proof I believe it to be quite possible. 

 There is one case, which I do not remember having seen re- 

 ferred to, in which nature has tried an experiment for us. I 

 was informed by the President of the Deaf-Mute College at 

 Washington that the male and female students frequently marry 

 after leaving the college, and that their children are rarely deaf- 

 mutes. But the point to which I wish to call attention is the 

 admitted fact that there is usually no disease or malformation of 

 the vocal organs in a deaf-mute. Now, before deaf-mutes were 

 taught to talk as they are now, they passed their whole lives 

 without using the complex muscles and motor-nerves by the 

 accurate coordination of which speech is effected. Here is a 

 case of complete disuse, and there must have been some conse- 

 quent atrophy. Yet it has, I believe, never been alleged that 

 the children of deaf-mutes exhibited any unusual difiicuity in 

 learning to speak, as they should do if the effects of disuse of 

 the organs of speech in their parents were inherited. Here is 

 at all events the material of an experiment ready to our hands. 

 An experiment to show whether the effects of use and disuse 

 were inherited might also be tried by bringing up a number of 

 dove-cot pigeons in a large area covered in with wire netting so 

 low as to prevent flight, at the same time encouraging running 

 by placing food always at the two extremities of the enclosure 

 only, or in some other way ensuring the greatest amount of 

 use of the legs. After two or three generations had been 

 brought up in this way, the latest might be turned ,'out among 

 other dove-cot pigeons, at the age when they would normally 

 begin to fly, and it would then be seen if the diminished wing- 

 power and increased leg-power of the parents were inherited. 



No doubt many better experiments might be suggested ; but 

 these are sufficient to indicate the character of such as do not 

 require that the offspring be submitted to the same conditions 

 as those which produced the change in the parents, and which 

 thus enable us to discriminate between effects due to inherit- 

 ance and those due to a direct effect of the conditions upon the 

 individual. The cases of the Saturnia and the shrimps are of 

 the latter kind, and in their very nature can afford no proof of 

 heredity. Alfred R. Wallace. 



The Conditions Determinative of Chemical Change : 

 Some Comments on Prof. Armstrong's Remarks. 



In a paper (Nature, vol. xlviii. p. 237, Proc. Chem. Soc. 

 1893, 145) bearing the above title, Prof. Armstrong discusses 

 the phenomena of contact action, particularly those of the kind 

 described by Mr. H. 13. Baker. The whole discussion appears 

 to us to be based on erroneous conceptions and to call for some 

 criticism, first, on the general position assumed by him and, 

 second, of the details which he brings forward to support that 

 position. 



Eight years ago Prof Armstrong defined chemical action 

 as "reversed electrolysis." It is not clear from his remarks 

 whether this is one of the views which recent observations have 

 led him to modify ; but, assuming that he still holds that belief, 

 it may be pointed out that it by no means follows that because 

 an electric current can eflTect a chemical change, every chemical 

 change is due to or accompanied by electric action. It might 

 as well be argued that because a stone let fall on a glass plate 

 can shiver it, a shivered plate glass always implies a falling 

 stone as its cause— it could be broken by irregular rise of tem- 

 perature, or by loading it with a too heavy weight, phenomena 

 which imply no expenditure of kinetic energy. Yet the state- 

 ment contains a germ of truth, but only when so qualified as to 

 amount to something very different. Electrical energy may be 

 absorbed in effecting chemical decomposition ; when chemical 

 combination occurs some form of energy is made manifest. 

 1 he lacts, apart from theory, as we know them, appear to be 

 these. A certain fraction of some definite amount of electrical 

 energy may be absorbed in producing chemical decomposition, 

 and that fraction will be quantitatively converted into chemical 

 energy ; the electrical energy disappears as such, and elements 

 may be liberated from a compound, containing, as elements, the 

 equivalent quantity of chemical energy. These elements may 

 part with their chemical energy, which will then cease to exist 



