414 



NA TURE 



[March 2, 1899 



of the whole book. It contains a full account of thernio- 

 chemical notation, a selection of the more important data, 

 and an elementary discussion of the law of maximum 

 work. Chemical equilibrium and dissociation are also 

 dealt with, the treatment being non-mathematical, and 

 bearing evidence of the influence of van 't HofT. The 

 book as a whole forms an admirable introduction to 

 general chemistry ; the student who has mastered its 

 contents will have nothing to unlearn, and will be able 

 to proceed at once to the larger textbooks of Ostwald 

 and van 't Hoft". 



Elementary Hydrostatics. By Charles Morgan, M.A., 

 R.N. Pp. 106. (London : Rivingtons, 1899.) 



Thls small text-book contains practically a condensed 

 account of all the leading points in hydrostatics which 

 are usually included in an elementary course, accom- 

 panied by an unusually large number of exercises. It 

 makes no attempt at exhaustive treatment, and is rather 

 intended for those studying the subject with tutorial aid. 

 We are sorry to see that the author has gone on the old 

 lines in the dual interpretation of "pressure" as thrust, 

 and also as thrust per unit area, and we should have 

 liked to have seen the notion of " whole pressure •' kept 

 in the background, and greater emphasis laid on the use 

 of the formula for the same in obtaining the resultant 

 thrust on a plane area. The familiar figure of the air- 

 condenser with the valves resting in their usual impossible 

 upside-down position is here once more reproduced. We 

 like the author's simple treatment of centres of pressure 

 as being instructive and useful to beginners, despite the 

 objections that mathematicians may raise against its 

 validity. For the points which we have criticised, the 

 fault probably lies not so much with the author as with 

 the examinations for which it is his purpose to prepare 

 candidates, and we think that the book will be of great 

 value to all students whose limited time prevents their 

 reading a large treatise. G. H. B. 



The Valley of Light. — Studies with Pen and Pencil 

 in the Vaudois Valleys of Piedmont. By W. 

 Basil Worsfold. (London: Niacmillan and Co , Ltd., 

 1899.) 

 An author adds to his difficulties by writing a book in 

 the form of letters, especially when he desires to com- 

 bine instruction with entertainment. .Mr. Worsfold 

 has not been more successful than others in overcoming 

 these, and we are not surprised that, as he admits, his 

 fair correspondent found his epistles " not very enter- 

 taining." in fact he does not add much to our know- 

 ledge of this district. Like his predecessors, he is almost 

 silent on its geology and botany, and devotes himself to 

 the history of the past persecutions and present for- 

 tunes of the Waldenses. The former subject is an 

 interesting but hardly a novel one; for it is treated pretty 

 fully in lieattie's " VValdenses " and Gilly's "Narrative." 

 The Waldenses, in fact, have already been the cause of not 

 a few books, if we include those in other tongues than 

 our own, and Mr. Worsfold's does little more than add 

 to their number. We doubt, indeed, whether the best 

 authorities would agree with him m tracing the 

 Waldenses back to early Christian settlements in these 

 valleys, or in the date (twelfth century) which he assigns 

 to the Nobia Lecjon. Nothing of special importance 

 seems to have happened in the Waldensian valleys 

 during the last half-century. Their worthy inhabitants 

 have prospered fairly and maintained their high char- 

 acter, but this, though satisfactory, affords but few 

 opportunities to an author. In short, Mr. Worsfold's 

 book has no scientific value, for even the illustrations 

 are poor ; and it displays little historical research or 

 originality. 



NO. I 53 I, VOL. 59] 



LETTERS TO THE EDITOR. 



[The Editor does not hold himself responsible for opinions ex- 

 pressed by his correspondenls. Neither (an he unierlake 

 to return^ or to correspond with the writers of, rejected 

 manuscripts intended for this or any other ^-art of Nature. 

 A'o notice is taken of anonymous communications.^ 



Earthquake Precursors. 



I.N Na 1 LKE, Febru.iry 16 (p. 36S), under the title of " Earth- 

 quake Echoes, " I described the more or less rhythmical series 

 of fading resultants which are seen to succeed many large earth- 

 quakes. These earthquake followers, the first of which may 

 sometimes represent the Vri Kaishi, or return shaking of 

 the Japanese, are repeated in groups with a decreasing ampli- 

 tude, an increasing period, and wilh a •smoother and smoother 

 contour. The last of the series may be so small that it .- 

 usually difficult to say with certainly when a large earlhqua'.v j 

 has ceased to exist. As pointed out by Mr. K. D. Oldham, 

 it seems quite possible that certain of the terminal vibrations 

 may have travelled round the world in a direction opposite to 

 that taken by the larger members uf the series. The move- 

 ments to which I now refer are the procession of vibrational 

 groups which run before the main disturbance, with the smaller 

 of which, under the name of preliminary tremors, we are al- 

 ready more or less familiar. These precursors have in several 

 respects characteristics which are exactly the op)x>sile to those 

 of the earthquake followers. They have a definite commence- 

 ment, and with large earthquakes group after group usually 

 increases suddenly in amplitude and period. 



Another feature of the precursors is that, whilst group after 

 group may grow larger, they become more and more larger 

 featured in their contours. The very first of (he preliminary 

 tremors have no frctillements, or have lost whatever they may 

 have had, whilst those which follow carry serrations which are 

 well marked. This observation, together wilh ihat of the growth 

 in amplitude, suggests the idea that the main features of each 

 group of precursors starling from a common origin have reached 

 an observing station by diflferenl routes ; the first have come 

 along Knott's path of least time, whilst ihe latter, culminating 

 in the shock, may have travelled along paths continually ap- 

 proximating to thai of a free surface-wave. 



Now and again, we see in groups of jjreliminary tremors a 

 likeness in contour and arrangement of what is to follow ; but 

 likenesses of this description are perhaps best seen when we 

 compare the shock and its immediate forerunners with the 

 Uri Kaishi, or first echo and its successors. Thus, in the 

 accompanying photographic reproduction of ihe disturbance of 



B 



•<m'CW, 



June 29, 1898, if we take S S as a line of symmetr)', which lines 

 are not uncommon in seismograms, the shock i in the group A 

 is preceded by groups of waves 2, 3, 4 and 5, which are not 

 unlike the echoes i', 2', 3', 4' and 5' ; whilst in the precursors 

 B and C, i, 2, 3 are not unlike I, 2, 3. In group D all like- 

 nesses are lost. Our knowledge of the very first preliminary 

 tremors like D is less than that of those which follow. Near 

 to an origin they may have a duration of from one or two up 

 to ten or twenty seconds, and their period has been recorded 

 at from 1/5 to 1/20 of a second. When iliey are preceded by 

 a sound-wave, we have evidence of a very much higher fre- 

 quency. If these vibrations have travelled long distances and 

 through our earth, most records indicate a jieriod of three or 

 four seconds. Records from Rome have shown periods of less 

 than half a second, but even these are probably much too large. 

 My own records indicate only a slight switching at the end of 

 a light elastic boom or a very rapid toandfro motion of the 

 boom relatively to its steady-point. Until a steady-point 

 seismograph with extremely light multiplying indices like that 

 of Viccnlini, or some other special form of apparatus, has been 

 employed as a recorder, our knowledge of this end of the seismic 

 spectrum is not likely to increase. 

 The last points connected with the earthquake precursors are 



