572 



NA TURE 



[October 15, 1896 



compound like NaCl, which is decomposed into two ions Na 



and CI, will thus produce double the normal effect on the 

 osmotic pressure and its conset|uences the diminution of vapour 

 ]>ressure and the lowering of the freezing jioinl. In the same 

 way, a molecule like H._,S04, which gives three ions, will 

 produce three times the effect which would be obtained if it 

 were undissociated. 



Thus Prof. Poynling's conditions would be satisfied, and at 

 the same time the advantages of the dissociation theory would 

 be retained. W. C. D. VVhetiiam. 



Trinity College, Cambridge, October 12. 



Responsibility in Science. 

 As one who supposes himself a physicist, I wish to protest 

 against some of Prof. Poulton's remarks in his recent British 

 Association address, as given in Nature, September 24. 



From the statements on p. 502, one would suppose that 

 physicists as a body had long been tyrannising over geologists 

 and zoologists, and that this reign of terror had remained 

 unbroken until recently, save for some slight diversions afforded 

 by mathematicians. 



When it comes to details, the physicists seem to resolve them- 

 selves into two individuals. Lord Kelvin and Prof. Tait, and 

 perhaps a third, von Helmholtz ; all of whom, by the way, 

 have an equally good claim to the title mathematician. Prof. 

 I'oulton apparently regards all physicists as committed to every 

 theory iirojjounded by every individual physicist. This would 

 certainly be unlimited liability with a vengeance. 



Personally I do not hold myself committed to the truth of any 

 theory, past, present, or future, until such time as I have 

 explicitly signified my assent to it. If one were explicitly to 

 signify one's dissent from every physical theory, or every state- 

 ment of jihysical facts, which one is not prepared to accept, 

 there might be little time left to do anything else. Perhaps I 

 can bring this home most clearly to Prof. Poulton by expressing 

 my views as to one or two of his own statements. 



On page 502 he says, " the earth, even when solid, will alter 

 its form when exposed for a long time to the action of great 

 forces" (italics mine). Here, and in the rest of the jmssage, is a 

 strong flavour of the erroneous view that a solid is rigid in the 

 mathematical sense, except when viscous under great and 

 ])rolonged stress. It is surely time that scientific men in all 

 departments grasped the conception of elastic strain and dis- 

 placement. 



On the same page are other imperfections in the statement of 

 the arguments against deducing the time of conscjlidation of the 

 earth from its present form. Prof. Poulton apparently considers 

 it proved that the earth's angular velocity of rotation is diminish- 

 ing, and that the only agent to be considered is the action of the 

 tides. If, however, the earth's temperature is diminishing, and 

 the material contracts in cooling — conclusions most generally 

 accepted— the consequent diminution in the moment of inertia 

 tends to shorten the period of rotation. Such shortening was in 

 fact made the basis of his speculations by the eminent French 

 mathematician Prof. E. Roche (Academic . . . de Montpellier, 

 Mt'moires de la Scetion des Sciences, vol. x., 1880-S4, p. 232). 



On page 503, we are told "there is some evidence which 

 indicates that the interior of the earth in all probability conducts 

 better than the surface. Its far higher density is consistent with 

 the belief that it is rich in metals, free or combined. Prof. 

 Schuster concludes that the internal electric conductivity must 

 be considerably greater than the external." 



When one considers the enormous pressures which existing 

 theories jioint to in the earth's interior, and remembers how 

 conspicuously less the accepted mean density is than that of the 

 lightest of the heavy metals under atmospheric pressure, one 

 can only recognise the inconclusiveness of the evidence from 

 this source. 



The reference to Prof. Schuster's conclusion is ambiguou.s. 

 Does Prof. Poulton believe electrical and thermal conductivity 

 necessarily to vary together ? If so, then the fact that electrical 

 conductivity diminishes in metals and ordinary alloys as the 

 temjierature rises, is one he ought to consider. In any case he 

 might be well advised to allow for the possibility that Lord 

 Kelvin's speculations do not possess a monopoly of physical 

 uncertainties. 



The direct experiments by Lord Kelvin (Nature, June 1895, 

 p. 182) on the influence of temperature on thermal conductivity 



NO. 1407, VOL. 54] 



are very probably, in Prof. Poulton's opinion, not sufliciently 

 varied, as regartis either material or range of temperature, to 

 form a .sub.stantial basis for wide conclusions. Still I .should 

 have expected him to refer to them, if only to mark his 

 recognition of an attem]it on Lord Kehin's part to meet his 

 critics with something better than .surmises. 



Our uncertainty as to the true value of the mean temperature 

 gradient near the earth's surface might fairly, I think, have been 

 adduced by Prof. Poulton. Observations have, in fact, been 

 limited to comparatively small areas of the .surface, and the 

 results obtained have varied much. There are also sources 01 

 error whose elimination is difficult. Irregularities in the surface, 

 the presence of the recording apjiaratus, and the disturbance 

 caused by previous excavations, tend to alter the temperature it 

 is intended to measure ; while the conditions may prejudice the 

 correct working of the apparatus. An instructive example of 

 this last defect came under my notice recently. Very fairly 

 accordant readings with two maximum thermonrcters in a deep 

 boring full of water indicated an excess of some'30' F. in the bottom 

 over the surface temperature. Direct experimeiU in a hydraulic 

 press proved, however, that fully half the rise was fictitious, 

 being simply due to the contraction of the bulbs under the 

 pressure to which they were exposed. C. Chrek. 



September 28. 



The Climate of Bremen in Relation to Sun-spots. 



May I invite attention to the variation of mean temperature 

 of the summer half (April to September) at Bremen, which 

 seems to me to suggest sun-spot influence ? The observations 

 used are those given by Dr. Bergholz in his " Ergeljnisse." 

 They extend (with a break from 1S14 to 1823) from the begin- 

 ning of the century. 



The dotted curve in the diagram (a) shows this variation. 

 After smoothing _with averages of 5, we have the continuous 



Jj 3o 3S 'M Aj So :ff Vp 'iyyo 'yS'FO Ss 'ff fi 



/Seo 0!,- 10 

 a. Dotted curve 



Vi' iO fi~ 60 



me.in temperature April to September, Hremen. Continuous 

 tiled with averages of 5. *. Sun-spot curve (inverted). 

 Smoothed curve of rainfall, August and September, Bremen. 



curve traversing the other. Below, (*), is the sun-spot cilrve, 

 inverted, and a general agreement will be made out with the aid 

 of the figures given ; the wave-crests of the smoothed curve 

 corresponding, more or less, with the minima of sun-spots. This 

 result is in harmony with those of Koppen and others. 



We may note the salient years (half-years) 1811, 1S34, 1842 

 (1846), 1857, 1868, 1878, 18S9 ; all near minima. 



Representing the half-years as -t- or - , according as they are 

 above or below the average, and selecting the sun-spot maximum 

 and minimum years, I find five out of seven of those seasons 

 in the former case (maximum) to be below average, and four out 



