38 



NATURE 



[March ii, 1920 



sun his time standard would not be that determined 

 by bodies he had carried with him, but the standard 

 found by observing from the sun similar bodies on 

 the earth, and he would judge that his time standards 

 were changed by being displaced. Of course, if they 

 were not changed, the spectral shift would be zero. 

 The colour analogy, however, shows that there is no 

 special reason to believe that they are unchanged, and 

 it certainly seems most likely that the invariable 

 quantity in such a displacement is As, for this is 

 already known to be of fundamental importance in 

 other problems. The shift, therefore, is probable, 

 though if it were absent it would not be very difficult 

 to construct a theory that would fit the fact. 



If it were true that it was the same for atoms on 

 the sun and on the earth, we might expect our 

 standards of length also to be the same; but this 

 leads to a surprising result, for if they were, the 

 measure of the wave-length of the emitted light would 

 be proportional to {git/gii)i, so that it would not be 

 possible to continue to use the wave-length as a 

 standard of length; thus such a hypothesis would 

 lead, not to a simplification, but to an added com- 

 plexitv. It may also be noted that the spectral shift 

 depends on the part of Einstein's law that agrees 

 with Newton's, so that the two stand or fall together 

 if this phenomenon is crucial. 



Einstein's law, however, rests on firmer ground 

 than the theory of the spectral shift. As to whether 

 this shift exists, the available data on an average point 

 to one of nearly the predicted amount, and are cer- 

 tainly much nearer this than zero. They show a great 

 variation in the amount of the shift, which must be 

 explained before the question can be regarded as 

 solved. Many causes are capable of producing this 

 variation, but what seems to me likely to be the chief 

 does not appear to have received much attention. The 

 prediction rests on the assumption that the vibrating 

 atoms are in similar surroundings, which is plainly 

 false. It is, indeed, required by the theory of stellar 

 evolution that the whole constitution of a star must 

 alter owing to successive types of atom becoming un- 

 stable and passing over into more stable forms. In- 

 stability demands that the slowest free vibration of 

 the atom has its frequency reduced to zero, and in the 

 process the other periods must be affected. The 

 remarkable fact is not that there are shifts, but that 

 the observed spectra are as much like terrestrial ones 

 as they are. 



Harold JeffrI'YS. 



Meteorological Office, S.W.y. 



The Position of the Meteorological Office. 



That the study of the atmosphere and the practical 

 applications of meteorology should be supported and 

 encouraged by the Government is a proposition so 

 obvious that it is accepted in every civilised country. 

 It does not, however, follow that the meteorological 

 service of a country should be conducted as a branch 

 of the civil service, still less of the military service, 

 and British meteorologists must be grateful for the 

 emphasis laid in the leader in Nature of February 26 

 on the importance of scientific control of official 

 meteorology. 



I do not know enough of the present constitution 

 of the Meteorological Office to offer any criticism of 

 the Air Ministry in relation to it, but I am verv 

 strongly in agreement with the resolution of the 

 Roval Meteorological Societv as to the importance of 

 full inquiry before changing the constitution of the 

 Meteorological Office, which has led to such remark- 

 able advances in meteorological science since 1905. 



NO. 2628, VOL. 105] 



The transfer to the Department of Scientific and 

 Industrial Research, which you state to have been 

 contemplated at one time by a . Committee of the 

 Cabinet, would, it seems to me, have been a natural 

 development of the constitution under the Meteoro- 

 logical Committee, and it would have been free from 

 the dangers to scientific progress which are, not un- 

 naturally, feared from a subordinate position in the 

 Air Ministry. Had a full inquiry been held, I doubt 

 whether the claims of the Air Ministry would have 

 been preferred to those of the Board of Agriculture 

 and Fisheries, the Admiralty, the Board of Trade, 

 and, in particular, to those of the Ministry of Health, 

 The union of the British Rainfall Organization with 

 the Meteorological Office has altered its centre of 

 gravity so far as to make its equilibrium less stable 

 in the Air Ministry than it would be in either the 

 Board of Agriculture and Fisheries or the Ministry 

 of Health. As part of the Department of Scientific 

 and Industrial Research the Meteorological Office 

 would be in neutral territory, and could be equally 

 serviceable to all the great Departments, each of which 

 would naturally be represented on the Advisory Com- 

 mittee controlling the organisation. The position 

 would then be analogous to that of the Geolof*ical 

 Survey, which, perhaps, is the official scientific body 

 most nearly akin to the Meteorological Office. 



For scientific bodies of this kind freedom from all 

 unnecessary trammels of officialdom is necessary in 

 order to permit the expansion and development which 

 are essential to healthy life and practical usefulness ; 

 and in a body of such universal usefulness as the 

 Meteorological Office in its present expanded form 

 some representation of the industrial and economic 

 applications of meteorology upon the advisory com- 

 mittee or other controlling board is nearly as im- 

 portant as the representation of independent men of 

 science. 



Hugh Robert Mill. 



Hill Crest, Dorman's Park, Surrey, 

 March 2. 



The issue of Nature for February 26 contained an 

 account of the Royal Meteorological Society's resolu- 

 tion in reference to the transfer of the Meteorological 

 Office to the Air Ministry, a leading article dealing 

 with the same subject, and correspondence on the 

 organisation of scientific work, part of which seems 

 directly applicable to the same theme. 



If it be true that the Meteorological Committee is no 

 longer to exist, the society's protest appears amplv 

 justified. Otherwise the position of the Meteorological 

 Office as a branch of the Air Ministry, with a scientific 

 advisory committee, would appear not verv dis- 

 similar to that of the Natural History Museum ; or 

 perhaps a better comparison would be with the Royal 

 Observatory, Greenwich, which is under the Ad- 

 miralty, the Astronomer Royal being supported by a 

 scientific advisory committee in the shape of the Board 

 of Visitors, of whom only one, the Hydrographer, 

 directly represents the Admiralty, the rest being either 

 university professors of astronomy or else expressly 

 nominated by the Royal Society or the Royal .Astro- 

 nomical Society. 



The Meteorological Department at Greenwich, 

 though now in its eightieth year, is too recent to 

 expect direct representation on the Board, especially 

 as its activities have not generally run In the direc- 

 tion of research, but the fact remains that the work 

 at Greenwich has points of contact not only with the 

 Admiralty, but also with the Board of Trade, the Post 

 Office, the Meteorological Office, the Colonial Office, 

 and other bodies. It oug^ht not to be Impossible to 



