TRANSACTION'S OF SECTION A. 519 



to detect varintions of 3" in the observations wliicli are taken to 6", with a 

 probability of error of only one in 2,000, is, I think, a proof of the value of the 

 method and the carefulness of the observations. The periodograph has another 

 valuable use. It not only jrives us the tiuie necessary to establish true periodicities 

 of given amplitude, but it also gives us an outside limit of the time beyond which 

 an accumulation of material is of no further advantage. That limit is reached 

 when the time is sufficient to discover the smallest amplitude which the instrument, 

 owing to its imperfections, will allow us to detect. 



I am only concerned to-day with a purely statistical inquiry, and not with the 

 explanation of any suggested relationship. To prevent misunderstandings, how- 

 ever, I may state "that I consider the possibility of a direct magnetic or electric 

 action of the moon excluded. As regards the latter the diurnal variations of 

 electric potential would be so strongly atfected by a lunar electrification sufficiently 

 strong to influence the outbreak of thunderstorms that it could not have escaped 

 discovery. We mu.'.t not, however, be dogmatic in asserting the impossibility of 

 indirect action. The unexpected discovery of radio-activity has opened out an 

 entirely new field, and we cannot dismiss without renewed careful inquiry the 

 evidence of lunar action which 1 have given. Its reality can be decided by obser- 

 vation only. No, not by observation only, but by observation supplemented by 

 intelligent discussion ; and this brings me to my concluding appeal, which I wish 

 to urge upon you with all the legitimate weight of strong conviction and all the 

 illegitimate influence of presidential infallibility. 



The subjects with which our sub-section is concerned deal with facts which are 

 revealed to us by observation more frequently than by experiment. There is, in 

 consequence, a very real danger that the importance of observation misleads us 

 into mistaking the means for the end, as if observation alone could add anything to 

 our knowledge. Observation is like the food supplied to the brain, and knowledge 

 only comes through the digestion of the food. An observation made for its own 

 sake, and not for some dehnite scientiflc object, is a useless observation. Science 

 i.s not a museum for the storage of disconnected facts and the amusement of the 

 collecting enthusiast. I dislike the name ' observatory ' for the astronomical work- 

 shop, for the same reason that I should dislike my body to be called a food 

 receptacle. Your observing dome would be useless without your computing room 

 and your study. What you want ^s an astronomical laboratory, a meteorological 

 or magnetic laboratory, attaching to the word ' laboratory ' its true meaning, 

 which is a workshop in which eyes and hands and brain unite in producing a 

 combined result. 



The problems which confront the astronomer being more delinite than those of 

 meteorology, astronomy has grown under the stimulus of a healthy tradition. 

 Hence it is generally recognised, at any rate in the principal observatories, that 

 the advance of knowledge is the chief function of the observer. Nevertheless the 

 President of the Astronomical Department of Section A last year (Professor H. H. 

 Turner) has found it necessary, in his admirable Address, to warn against the 

 danger there is that the astronomer should allow himself to be swallowed up in 

 routine work and mere drudgery. 



The descent is easy: you begin by being a scientific man, you become an 

 observer, then a machine, and finally — if all goes well — you design a new eyepiece. 



If such a danger exists in astronomy, what shall we say about meteorology ; 

 That science is bred on routine, and drudgery is often its highest ambition. The 

 heavens may fall in, but the wet bulb must be read. Observations are essential, 

 but though you may never be able to observe enough, I think you can observe too 

 much. I do not forget the advances which meteorology has made in recent years, 

 but if you look at these advances 1 think you will find that most of them do not 

 depend on the accumulation of a vast quantity of material. The progress in some 

 cases has come through theory, as in the applications of thermodynamics or 

 through special experiments, as by kite and balloon observations, and when it has 

 come through the ordinary channels of observation only a comparatively short 

 period of time has been utilised. It woidd not be a great exaggeration to say that 

 meteorology has advanced in spite of the observations and not because of them. 



