December 17, 1908] 



NATURE 



185 



in foni! and more than usual correctness, and fre- 

 quently in an unusual point of view. This is parti- 

 cularly noticeable in the chapters devoted to seismo- 

 graphs, which are refreshing in the absence of any 

 Dolemical advocacy of one pattern of instrument or 

 depreciation of another ; there is little in the way 

 of description of particular instruments or types of 

 seisniDgraph, and no attention is devoted to details 

 of mechanical construction, which may vary accord- 

 ing to the purpose of the instrument, but instead 

 we have an impartial statement of the principles 

 on which their construction is based and which 

 control their action. The dynamics of the horizontal 

 pendulum, which have been the subject of both 

 mathematical and experimental investigation, are 

 treated in a manner which makes them clear to any- 

 one able to follow the simple mathematics used in the 

 text, but it is unfortunate that Dr. Knott had not 

 more mercy on those less mathematically disposed 

 than himself, and expressed his numerical results in 

 a form more immediately intelligible than that adopted 

 by him. 



This question of the behaviour of the horizontal 

 pendulum in response to a periodic undulatory tilting, 

 as opposed to its response to a static tilt, is one which 

 has an important bearing on the design of seismo- 

 graphs; in most of these the design has been to 

 eliminate resistance so far as possible, but there is 

 another school which deliberately introduces a damp- 

 ing device of sufficient power to make the pendulum 

 dead-beat or aperiodic, and it has been claimed that this 

 damping renders the record accurate and capable of 

 interpretation in terms of the displacement produced 

 by a static tilt. Dr. Knott's figures show that this 

 claim is unfounded. Where the period of the undula- 

 tion is not less than three times that of the free swing 

 of the pendulum, the amplitude of the record is within 

 10 per cent, of the displacement due to a static tilt of the 

 same angle, the error being in excess in the case of the 

 free and in defect in the case of the damped pendulum. 

 \\'hen the period of the undulation approaches nearer 

 to equality with that of the pendulum, the amplitude 

 of the record increases largely in the case of the un- 

 damped pendulum and becomes diminished in the case 

 of the damped pendulum, but in neither tvpe is it 

 possible to determine the true value of the angular tilt 

 from the amplitude of the record. From this it will 

 be seen that the result of a complete damping of the 

 pendular swing is a diminution of sensitiveness of 

 the instrument, and as it is onlv when the period of the 

 undulation reaches three times that of the pendulum 

 that either form gives a record capable of approxi- 

 mate interpretation in terms of the static lilt, there is 

 no material difference in accuracy between the two 

 when this limit is reached. 



The periodicity of earthquakes is discussed at some 

 length, with the general result that there is little 

 evidence of the reality of any of the periods believed 

 to have been established. We are not only in coin- 

 l)Iete agreement with this conclusion, but would go 

 even further than Dr. Knott in our distrust of the 

 ulilii\- (jf applying the method of harmonic analysis 

 lo the discussion of effects the caus?s of which do not 

 NO. 2042, VOL. 7g] 



vary in a harmonic manner, and the method seems 

 particularly inapplicable to the discussion of the effect 

 of tide-producing stresses in the causation of earth- 

 quakes. The amount and direction of this stress, at 

 any given instant and place, depend on the zenith 

 distance, not on the hour angle, of the tide-producing 

 body, and though these vary with each other, they 

 do not vary in any uniform proportion. In these cir- 

 cumstances an harmonic analysis of the time of occur- 

 rence of earthquakes seems calculated to obscure rather 

 than elucidate any direct effect of the tide-producing 

 force, though it might reveal a tidal effect of a 

 different nature. 



For the rest the book is an adequate and clearly 

 expressed treatment of the subject it professes to deal 

 with. It cannot be described as easv reading, yet the 

 difficulty lies entirely in the accuracy of its expression, 

 and the consequent necessity for the frequent use of 

 words unfamiliar except to the trained phvsicist, but 

 anyone w-ho is desirous of understanding', and will 

 take the trouble to master the meaning of these un- 

 familiar terms, will find no difficulty in following the 

 argument. 



METHODS OF ACCURATE CALORIMETRY. 

 Methodes de CalorimHrie usities au Laboratoire 

 thcrmique de I'Uiiiversite de Moscoii. By Profs. 

 \\'. Louguinine and .\. Schukarew. Translated 

 from the Russian by G. T. Gazarian. Pp. iii+192. 

 (Paris: A. Hermann; Geneve: Georg et Cie., 

 iqoS.) Price 8 francs. 



THIS volume by the well-known director of the 

 thermal laboratory at Moscow University and 

 his chief of staff does not claim to be a compre- 

 hensive treatise on all branches of calorimetric work, 

 but, nevertheless, it will be welcomed as placing 

 before a wider public the results of much valuable 

 research hitherto comparativelv unknown,_^ especially 

 in detail. Some of Prof. Louguinine's ingenious de- 

 vices for carrying out accurate calorimetric investiga- 

 tions have been partly described in specialist 

 treatises, but we have here complete descriptions, 

 with full and clear working drawings, published, we 

 believe, for the first time, except in their original 

 Russian. 



In calorimetry, perhaps to a greater extent than 

 in most branches of physics, very much of the success 

 attained in a particular experiment depends on atten- 

 tion to what might be considered small details. In 

 our opinion, one of the most valuable features of 

 the book is the large number of " wrinkles " or 

 " tips " given by the writers from their own experi- 

 ence on just those points on which the ordinary books 

 are silent. 



The first chapter is an excellent discussion of the 

 various tvpes of thermometers used in calorimetry. 

 The writers point out the absurdity of adhering to the 

 German form of thermometer with milk-glass scale, 

 carrying the graduations behind a thin capillary tube 

 and enclosed in an outer sheath. Even if the milk- 

 glass scale is fastened more or less by fusion at one 

 or the other end of the tube, the type has many draw- 

 backs, and would probably have been replaced long 



