450 



NA TURE 



[September 8, i\ 



gambler take some credit to liimself for the first suggestion of 

 the method of least squares, and the first discussion of the 

 integration of partial differential equations with finite differences 

 contained in Laplace's famous " Theorie Analytique des 

 Probabilites " ? 



The question asked Rankine by James R. Napier regarding 

 the horse povver which would be necessary to propel, at a given 

 rate, a vessel which Napier was about to build, resulted in the 

 many theoretical investigations carried out by Rankine on water 

 lines, skin-friction, stream lines, cS:c. For, as Prof. Tait has 

 said, "Rankine, by his education as a practical engineer, was 

 eminently qualified to recognise the problems of which the solu- 

 tion is required in practice; but the large scope of his mind 

 would not allow him to be content with giving merely the solu- 

 tion of those particular cases which most frequently occur in 

 engineering as we now know it. His method invariably is to 

 state the problem in a very general form, find the solution, and 

 apply this solution to special cases." 



Helmholtz studied physiology because he desired to be a 

 doctor, then physics because he found that he needed it for 

 attacking physiological problems, and lastly mathematics as an 

 aid to physical research. But I need not remind you that it is 

 his splendid work in mathematics, physics, and physiology, and 

 not his success in ministering to the sick, that has rendered 

 his name immortal. 



Did not Kepler ask : " How many would be able to make 

 astronomy their business if men did not cherish the hope of 

 reading the future in the skies?" And did he not warn those 

 who objected to the degradation of mingling astrology with 

 astronomy, to beware of " throwing away the child with the 

 dirty water of its bath"? Even now, may we not consider all 

 the astronomical research work done at the Royal Observatory, 

 Greenwich, as a bye-product, since the Observatory is officially 

 maintained merely for the purposes of navigation ? And are 

 there not many of us who feel assured that, since researches 

 in pure physics and the elucidation of new physical facts must 

 quite legitimately spring from routine standardising work, the 

 most direct way — even now at the end of the -nineteenth cen- 

 <tury of securing for the country a National Physical Laboratory 

 •is to speed forward a Government standardising institute ? 



Lastly, as you will find in Dr. Thorpe's fascinating " Life of 

 Davy," it was the attempt to discover the medicinal effect of 

 gases at the Pneumatic Institution in this city that opened up to 

 Davy the charm of scientific research. And, indeed, the Royal 

 Institution itself, the scientific home of Davy, Faraday, Tyndall, 

 Rayleigh and Dewar, owes its origin to Romford's proposal 

 " for forming in London by private subscription an establishment 

 for feeding the poor and giving them useful employment .... 

 ■connected with an institution for introducing and bringing 

 forward into general use new inventions and improvements by 

 which domestic comfort and economy may be promoted." 



Coming now to physics proper, there is one branch which, 

 although of deep interest, has hitherto been much neglected. 

 We possess three senses which enable us to detect the presence 

 of things at a distance — viz., seeing, hearing, and smelling. 

 The first two are highly cultivated in man, and, probably for 

 that reason, the laws of the propagation of the disturbances 

 which affect the eyes and the ears have been the subject of much 

 investigation, whereas, although to many animals the sense of 

 smell is of far greater importance than those of seeing or 

 hearing, and although, even in the human brain, a whole seg- 

 ment — a small one in modern man, it is true — is devoted to the 

 olfactory fibres, the laws of the production and propagation of 

 smell have received practically no attention from the physicists. 

 For some time past it has, therefore, seemed to me to be of 

 theoretical and practical importance to examine more fully into 

 the 'physics of smell. Various other occupations have hitherto 

 prevented my advancing much beyond the threshold of the sub- 

 ject, but, as it seems to me to open up what is practically a new 

 faeld of inquiry for the physicist, I take this oportunity of put- 

 ting on record some facts that have been already elucidated. 



Various odoriferous substances have been employed in the 

 experiments, and for several of these I am indebted to Mr. W. 

 J. Pope. Although the physicist has been allowing the 

 mechanical side of the subject to lie dormant, the chemist, I 

 find, has been analysing flowers and other bodies used in the 

 manufacture of scents, and then synthetically preparing the 

 odoriferous constituents. In this way, Mr. Pope informs me, 

 there has been added to the list of manufactured articles, during 

 the past seven years or so, vanilin, heliotropin, artificial musk, 



NO. 1506, VOL. 58] 



irone and ionone, wh'ch give the perfume: of the violet ; citral, 

 that of lemongrass ; coumarin, that of hay, and various others ; 

 and specimens of several of these artificial scents, together 

 with other strongly- smelling substances, he has kindly furnished 

 me with. 



If it be a proof of civilisation to retain but a remnant of a 

 sense which is so keen in many types of dogs, then I may pride 

 myself on ha/ing reached a very high state of civilisation. But 

 with the present investigation in view, this pride has been of a 

 very empty character, since I have been compelled to reject my 

 own nose as quite lacking the sensitiveness that should char- 

 acterise a philosophical measuring instrument. The ladies of 

 my family, on the contrary, possess a nasal quickness which 

 formerly seemed to me to be rather of the nature of a defect, 

 since, at any rate in towns, there are so many more disagreeable 

 odours than attractive ones. But on the present occasion their 

 power of detecting slight smells, and the repugnance which 

 they show in the case of so many of them, have stood me- 

 in good stead, and made it possible to put before you the fol- 

 lowing modest contribution to the subject. 



There is a generally accepted idea that metals have smells, 

 since if you take up a piece of metal at random, or a coin out 

 of your pocket, a smell can generally be detected. But I find 

 that, as commercial aluminium, brass, bronze, copper, German- 

 silver, gold, iron, silver, phosphor-bronze, steel, tin and zinc 

 are more and more carefully cleaned, they become more and 

 more alike in emitting «^ smell, and, indeed, when they are z'^rj 

 clean it seems impossible with the nose, even if it be a 

 good one, to distinguish any one of these metals from the rest, 

 or even to detect its presence. Brass, iron, and steel are the 

 last to lose their characteristic odour with cleaning, and for some 

 time I was not sure whether the last two could be rendered 

 absolutely odourless, in consequence of the difficulty of placing 

 them close to the nose without breathing on them, which, as 

 explained later on, evolves the characteristic "copper" and 

 " iron " smell. But experiment shows that, when very con- 

 siderable care is taken both in the cleaning and the smelling, no 

 odour can be detected even with iron or steel. 



Contrary, then, to what is usually believed-, metals appear to 

 have no smell /^r se. Why, then, do several of them generally 

 possess smells ? The answer is simple ; for I find that handling 

 a piece of metal is one of the most efficient ways of causing it to 

 acquire its characteristic smell, so that the mere fact of lifting 

 up a piece of brass or iron to smell it may cause it to apparently 

 acquire a metallic odour, even if it had none before. This 

 experiment may be easily tried thus : — Clean a penny very care- 

 fully until all sense of odour is gone ; then hold it in the hand 

 for a few seconds, and it will smell — of copper, as we usually 

 say. Leave it for a short time on a clean piece of paper, and 

 it will be found that the metallic smell has entirely disappeared, 

 or at any rate, is not as strong as the smell of the paper on 

 which it rests. The smell produced by the contact of the 

 hand with the bronze will be marked if the closed hand con- 

 taining it be opened sufficiently for the nose to be inserted, and 

 it can be still further increased by rubbing the coin between the 

 fingers. 



All the metals enumerated above, with the exception of gold 

 and silver, can be made to produce a smell when thus treated, 

 but the smells evolved by the various metals are quite different. 

 Aluminium, tin, and zinc, I find, smell much the same when 

 rubbed with the fingers, the odour, however, being quite 

 different from that produced by brass, bronze, copper, German- 

 silver, and phosphor-bronze, which all give the characteristic 

 "copper" smell. Iron and steel give the strong "iron" 

 smell, which, again, is quite different from that evolved by the 

 other metals. In making these experiments it is important to 

 carefully wash the hands after touching each metal to free them 

 from the odour of that metal. It is also necessary to wait for a 

 short time on each occasion after drying the hands, since it is 

 not until they become again moist with perspiration that they 

 are operative in bringing out the so-called smells of metals. 



That the hands, when comparatively, dry, do not bring out 

 the smell of metals is in itself a disproof of the current idea 

 that metals acquire a smell when slightly warmed. And 

 this I have further tested by heating up specimens of all 

 the above-mentioned metals to 120° Fahrenheit, in the sun, and 

 finding that they acquire no smell when quite clean and 

 untouched with the hands. ^ 



Again, dealing with the copper group, or with aluminium, no 

 smell is produced by rubbing any one of them with dry table- 



