»92 



NATURE 



[August 31, 191 1 



Bradley's work ; he discovered aberration, not by any help 

 from Newton, but by accumulating a mass of observations. 

 He had no ready-made hypothesis, or rather he had a 

 wrong one, viz. that the stars would show displacement 

 due to parallax ; and after this was proved wrong, as it 

 was at the very outset, he had nothing in the way of a 

 theory to guide him, and found great difficulty in devising 

 one after he had collected his facts, which spoke for them- 

 selves so far as to reveal plainly the essential features of 

 the phenomenon in question. 



" Modern discoveries " (on the preceding page of the 

 " B. of P.") " have not been made by large collections of 

 facts, with subsequent discussion, separation, and resulting 

 deduction of a truth thus rendered perceptible." 



To this I venture to oppose not only such work as that 

 of Bradley, but much in the recent history of astronomy ; 

 the discoveries about systematic proper motions, about 

 moving clusters, about the growth of velocity with life- 

 history, and so forth. 



" There is an attempt at induction going on, which has 

 yielded little or no fruit, the observations made in the 

 meteorological observatories. The attempt is carried on in 

 a manner which would have caused Bacon to dance for 

 joy. . . . And what has come of it? Nothing, says M. 

 Biot, and nothing will ever come of it : the veteran mathe- 

 matician and experimental philosopher declares, as does 

 Mr. Ellis, that no single branch of science has ever been 

 fruitfully explored in this way." 



De Morgan was a mathematician, and I have noticed that 

 mathematicians are apt to be crisp in their statements : 

 but he is a bold man who says " nothing will ever come 

 of it." Perhaps an equally crisp statement on the other 

 side may be pardoned. I adventure the remark that if 

 nothing has hitherto come of such observations, it is 

 because observers have been misled by the very teaching 

 of De Morgan and others who share his views : they have 

 been told that they will do no good without a theory until 

 they have come to believe it ; whereas the truth probably 

 lies in a quite different direction. To present my reasons 

 for this proposition I must ask you first to consider in 

 some detail the method of discussing meteorological 

 observations suggested some years ago by Prof. Schuster. 

 He gave an account of it to the Department of Cosmical 

 Physics over which he presided in igo2, so that I must 

 face some repetition of what he said ; but the matter is 

 so important that I trust this may be pardoned. 



Let us compare the records produced on a gramophone 

 disc by the playing of a single instrument and by that of 

 an orchestra. The first will be comparatively simple, and 

 when suitably magnified will show a series of waves which 

 in certain parts of the record form sequences of great 

 regularity. These represent occasions when the single 

 instrument played a long sustained note, the pitch of which 

 is indicated by the frequency of the wave. If the instru- 

 ment plays more loudly, while still keeping to the same 

 note, the heights of the waves will increase, though their 

 frequency will not be altered. The exact shape of each 

 wave will represent the quality of tone which characterises 

 the instrument : and if another instrument were to play 

 the same note it would be different. But so long as we 

 keep to the same instrument, whenever the same note 

 recurred we should find, generally speaking, the same shape 

 of wave : and we could resolve it into its constituents, one 

 being the main wave and others harmonics of different 

 intensities. The analysis of such a record would thus be 

 a comparatively simple matter, on which we need scarcely 

 dwell further. Very different is the ens., of the orchestral 

 record. There are numerous instruments, playing notes of 

 different pitch, intensity, and character, each of which, if 

 playing alone, would produce its own peculiar record. But 

 when they play together the records are all combined into 

 one. The needle can only make one record, but it is a 

 true sum of all the individuals ; for when the instrument 

 is set to reproduce the playing of the orchestra, a trained 

 ear can perceive the playing of the separate instruments — 

 when the strings are playing alone, and when the wind 

 joins them: when the horn comes in, and whether there 



.' players or only one: nay, even that one of the j 

 second violins is playing somewhat flat ! This could not 

 happen unless the individual performances were essentially I 

 and truly existent in the combined record ; and vet this ) 



consists of only one single wavy line. The waves are, 

 however, now of great complexity, and it seems at fust 

 sight hopeless to analyse them. The mathematician knows, 

 however, that such analysis is possible, and is quite simple 

 in conception, though it may be laborious in execution. 

 Selecting a note of any given pitch, a simple calculation 

 devised by Fourier will reveal when and how loudly that 

 particular note was being played. This being so, it is only 

 necessary to repeat the process for notes of different pitch. 

 But though this can be stated so simply, the carrying out 

 in practice may involve immense labour, by reason of the 

 number of separate notes to be investigated. It is not 

 merely that these will extend from low growls by the 

 double bass to high squeaks by the fiddles, but that their 

 variety within these wide limits will be so great. The 

 series is really infinite. We might, indeed, prescribe a 

 certain scale of finite intervals for the main notes, as in 

 a piano : but the harmonics of the main tones would refuse 

 to obey this artificial arrangement, and would form inter- 

 mediate pitches, which must be properly investigated if our 

 analysis is to be complete. Moreover, the orchestral instru- 

 ments will not keep to any such prescribed intervals, but 

 will insist on departing from them more or less, according 

 to the skill of the performer. There is a story told of an 

 accompanist who vainly tried to adjust the key of his 

 accompaniment to the erratic voice of a singer. At length, 

 in exasperation, he addressed him as follows : " Sir, I have 

 tried you on the white notes, and I have tried you on the 

 black notes, and I have tried you on white and black 

 mixed : you are singing on the cracks ! " Some instru- 

 ments will almost certainly "sing on the cracks," so that 

 we shall not easily escape from the examination of a very 

 large number of possibilities indeed — we may well call 

 them all the possibilities within the limits of audibility. 

 The illustration is already sufficiently developed for pro- 

 visional use. My suggestion is that science has only dealt 

 so far with the easy records, and that the genuine hard 

 work is to come. If we can imagine a number of deaf 

 persons turned loose among a miscellaneous collection of 

 gramophone records, with instructions to make what they 

 could of them, we can readily imagine that they would 

 pick out those of single instruments first. We must make 

 the researchers deaf, so that they may not use the beau- 

 tiful mechanism of the human ear, which has as yet no 

 analogue in scientific work. Possibly something corre- 

 sponding to this wonderful and still mysterious mechanism 

 may ultimately be devised, and then the course of scientific 

 research may be fundamentally altered : but for the present 

 we must regard ourselves as deaf, and as condemned to 

 work by patient analysis of the records. It is perfectly 

 natural, and even desirable, to begin with the easy ones ; 

 and the finding of an easy one would no doubt in our 

 hypothetical case be a sensational event, reflecting credit 

 on the lucky discoverer, who would be hailed as having 

 detected a new law, i.e. a new simple case. But sooner 

 or later these will be used up, and we must attack the 

 more complex orchestral records in earnest. Shall we find 

 that the best music is still to come, as our illustration 

 suggests? 



But we must return to Prof. Schuster's suggested plan 

 of work. It is closely similar to that already sketched 

 for dealing with a complex gramophone record. Let us 

 consider the record of any meteorological element, such 

 as temperature or rainfall. When these records are put in 

 the form of a diagram in the familiar way we get a wavy 

 line, which has much in common with that traced by a 

 gramophone needle on a smaller scale. The sight of the 

 complexities is almost paralysing, especially when those 

 who would otherwise attack the problem are deterred by 

 the emphatic assertion that it is useless to do so without 

 the equipment of some guiding hypothesis. Most of the 

 obvious hypotheses have, of course, already been tried, 

 and the majority of them have failed. It is to Prof. 

 Schuster that we owe the vitally important advice to dis- 

 regard hypotheses and make a complete analysis of the 

 record. Of course the labour is great, but the genuine 

 is not afraid of labour: he has a right to ask, 

 of course, that it shall not be interminable ; and when 

 we are told that we must examine an almost infinite series 

 of possibilities, there would seem to be some danger of 

 tiiis. But in practice the work always resolves itself into 



NO. 



2183, VOL. 87] 



