August 31, 191 1] 



NATURE 



he has made. Even if the like has never happened before, 

 this scientific advance is at any rate due to little more 

 than the accumulation of facts which arranged themselves, 

 as Bacon hoped would naturally happen. But does it 

 detract from the merits of this fine piece of observational 

 work that it was suggested by no leading theory ? And 

 1 will ask even further : Would its merits have been less 

 if no such immediate induction had presented itself? To 

 this second question I can scarcely expect a general answer 

 in the affirmative ; it is so natural to judge by results, and 

 so difficult to look beyond them to the merits of the work 

 itself, that I .shall not easily carry others with me in 

 claiming that the merits of the observe] shall be assessed 

 independently of his results. And yet I affirm unhesita- 

 tingly that until this attitude is reached, we cannot do 

 justice to the observer. I believe it will be reached in the 



nd I shall endeavour to give reasons for this fore- 

 cast ; but I admit frankly that our habit of judging by 

 results will be hard to break, ft extends even to the 

 observer himself, and leads to the withholding of his 



ions from publication, so that he may himself 

 extract the results from them. In the pure interests of 

 the advance of knowledge, it would be far better to publish 

 the material, so that many brains rather than one might 

 work upon it. But the observer knows that by this course 

 he risks losing almost the whole value of his patient work, 

 which would pass as unearned increment to the particular 

 person who was lucky enough to make the induction. 

 Hence arise quarrels such as those between Flamsteed and 

 Newton ; the former refusing to publish his observations 

 until he had himself had an opportunity of discussing 

 them, while Newton and Halley exerted their powerful 

 influence in the contrary sense. This situation by no 

 means belongs to a bygone age ; it may and does arise 

 to-day, and will continue to arise so long as the recogni- 

 tion of the observer's work is inadequate. It was men- 

 tioned a few minutes ago that Mr. Campbell had incurred 

 adverse criticism by accumulating a considerable mass of 

 unpublished observations. Let me be careful not to 

 suggest that his primary motive was the desire to have 

 the first use of them, for I happen to know that there was 

 at least one other good and sufficient reason for his action 

 in the difficulty of finding funds for publication, a difficulty 

 with which observers are only too familiar. But, what- 

 ever the reason, there were those who regretted the delay 

 in publication as hindering the advance of science. The 

 whole question is a delicate one, and might have been 

 better left unraised at the moment but for a most curious 

 sequel, which puts clearly in evidence the importance of 



rver and the desirability of allowing him to dis- 

 cuss his own work. To make this clear, a small 

 digression is necessary. 



During the last half-dozen years astronomers have been 

 startled on several occasions by pieces of news of a par- 

 ticular kind, indicating the association of large, widely 

 scattered groups of stars in a common movement. The 

 discussion of these movements is to occupy the special 

 attention of this Section at one of our meetings, which is 

 an additional reason for brevity in the present allusion. 

 Possibly, also, most members of the Section have already 

 heard of Prof. Kapteyn's division of the great mass of 

 bright stars into two distinct groups flying one through 

 the other ; and, again, of the discovery by Prof. Boss of 

 a special cluster of stars in the constellation Taurus, 

 moving in parallel lines like a flock of migrating birds. 

 The fascination of this latter discovery, and of one or 

 two others like it, is that when the information supplied 

 by the spectroscope is combined with that furnished bv the 

 long watching of patient observers, we can determine the 

 distance of the cluster and its shape and dimensions. We 

 realise, for instance, that there is a large flat cluster 

 migrating just over our heads, so that one member of it 

 (Sirius) is close to our Sun — that is to say, only three or 

 four light-years from him. " Close " is a relative term ; 

 and the distance travelled by light in three years is from 

 some standpoints by no means despicable. But it is small 

 in comparison with the dimensions of the cluster, which 

 is about one hundred light-years from end to end. The 

 study of these clusters will doubtless occupv our close 

 attention in the immediate future ; and it is very natural 

 that the discovery of one should lead to the search for 



others. Accordingly, we heard last autumn with the 

 deepest interest, but with modified surprise, the announce- 

 ment of common movement in a class of stars of a par- 

 ticular spectral type. The announcement rested to some 

 extent on the work done at the Lick Observatory, much 

 of which has been published in an abbreviated form. But 

 .Mr. Campbell, in the Lick Observatory " Bulletin " already 

 quoted, gives reasons why he cannot accept the con- 

 clusion, which is vitiated, in his opinion, by the existence 

 of a systematic error in the observations. Now on such 

 a point as this the observer himself is at any rate entitled 

 to a hearing, and is often the best judge. To take proper 

 precautions against systematic errors is the business of the 

 observer, and his efficiency may very well be estimated by 

 his success in this direction — this would be a far safer 

 guide than to judge by results. But sometimes such errors, 

 which are very elusive, do not suggest themselves until 

 the observations have been completed, and must be 

 detected from the observations themselves. This, again, is 

 rightly the business of the observer, and the desire to free 

 his observations from such error is a perfectly sound and 

 scientific reason for withholding publication. In the pre- 

 sent instance the error is a peculiarly insidious one ; and, 

 indeed, we are not even certain that it is an error. It is 

 a possible alternative interpretation of the facts that the 

 stars with Class B spectrum are in general moving out- 

 wards from the Sun, and the additional fact that there is 

 a comparatively large volume of space round the Sun at 

 present empty of B stars would seem to favour this 

 alternative. But, as already mentioned, the observer him- 

 self prefers rather to credit his observations with systematic 

 error, which gives a spurious velocity of 5 km. per second 

 to stars of this type. Now it will readily be understood 

 how an error of this kind may appear doubled : two 

 vehicles travelling in opposite directions approach or recede 

 from one another with double the speed of either, and if 

 one were erroneously supposed to be at rest, the other 

 would be judged to travel twice as fast. In this way the 

 B stars in a particular portion of the sky were judged to 

 be travelling with a common motion of 10 km. per second, 

 which would have been a discovery of far-reaching import- 

 ance if true, but which the observer relegates to the cate- 

 gory of systematic errors. 



The illustration will suffice to remind us that the work 

 of the observer is far from being merely mechanical : it 

 demands also skill and judgment — skill in defeating 

 systematic error, and a fine judgment, born of experience, 

 of the success attained. All this is independent of the 

 generalisations which may or may not be arrived at. 

 Bradley's skill as an observer enabled him to discover the 

 Aberration of Light and the Nutation of the Earth's Axis ; 

 it was enhanced rather than lessened when he went on to 

 make further observations which, had he lived, would 

 have conducted him to the discovery of the Variation of 

 Latitude. After his death the world waited more than a 

 century for this discovery to be made; but Mr. Chandler, 

 who played a leading part in it, has declared that Bradley 

 was almost certainly on its track. It would almost seem 

 that an observer is only properly appreciated by another 

 observer. There are doubtless many who, assisted by the 

 knowledge that Bradley's skill had twice previously con- 

 ducted him to a discovery, would be ready to admit the 

 value of his later work, although he did not live to crown 

 it ; but how many of these could properly appreciate 

 Bradley without such assistance? 



I venture to think that the great brilliance of Newton 

 has dazzled our vision so that we do not see some things 

 quite clearly. 



" Had it not been for Newton," writes De Morgan in 

 his "Budget of Paradoxes," p. 56, "the whole dynasty 

 of Greenwich astronomers, from Flamsteed of happv 

 memory, to Airy, whom Heaven preserve, might have 

 worked away at nightly observation and daily reduction 

 without any remarkable result : looking forward, as to a 

 millennium, to the time when any man of moderate 

 intelligence was to see the whole explanation. What are 

 large collections of facts for? To make theories from, 

 says Bacon ; to try ready-made theories by, says the history 

 of discovery; it's all the same, says the idolater; non- 

 sense, say we ! ' ' 



But nothing of this will fit in with what we know of 



NO. 2183, VOL. 87] 



