April 1 8, 18S9] 



NA TURE 



583 



important of these conditions I take to be the equatorial 

 position of the Anuvhimi basin ; ihe second that it is situated 

 in the heart of the continent. Both of these, but more particu- 

 larly the former, determine it ,ts the seat of ascending air- 

 currents, and therefore of their dynamic cooling on a gigantic 

 scale, and it is to this dynamic cooling that the high rainfall of 

 the region is to be ascribed. 



Very probalily a considerable portion of the precipitated 

 moisture is locally re-evaporated, so that, as suggested long 

 ago by Sir J 'hn Herschel in the case of the Hrazilian forest 

 rainfall, the same water is precipitated again and again. There 

 are not, I believe, in the lower atmosphere, any steady winds 

 blowing outwards to cany away the evaporation of the damp 

 forest tract, and the main loss of waier to be supplied by 

 easterly or other winds is that carried off by the river drainage, 

 probably less than half of the rainfall. The air which has 

 ascended to the higher regions of the atmosphere as a i^art of 

 the main circulation of the globe, parts with nearly the whole 

 of its vnpour in the act of ascending. 



We have a case in some respects analogous to that of the Upper 

 Aruwhimi in the very damp and equally forest-clad province of 

 Upper Assam. This too is characterized by a very calm atmo- 

 sphere, being girt with lofty mountains on the north and east, 

 and also shut off" on the south and south-west from the Bay of 

 Bengal by hills of considerable elevation. Such gentle winds as 

 blow in the valley are chiefly from the east or down valley. Yet 

 the rain'^all is over 100 inches in the year, and the whole tract is 

 one of marsh and dense forest. It isindeed not situated underihe 

 equator, and herein it is less favourably conditioned as a region 



Icessive rainfall than the basin of the Aruwhimi. 

 the result of a long study of the rainfall of India, and 

 ips no country affords greater advantages for the purpose, I 

 become convinced that dynamic cooling, if not the sole 

 of rain, is at all events the only cause of any importance, 

 fiat all the other causes so frequently appealed to in popular 

 lure on the subject, such as the intermingling of warm and 

 cold air, contact with cold mountain slopes, &c., are either 

 inoperative or relatively insignificant. 



Folkestone, April 11. Henry F. Blanford. 



" Les Tremblements de Terre." 



M. FouQU^'s letter (Nature, March 28, p. 510) does not 

 ijieet the main points of my criticism of his book. He thinks 

 that a pendulum swinging in synchronism with the ground's 

 motion is the right thing to use as an absolute seismometer. 

 M. Poincare's mathematical note, to which he refers as support- 

 ing his view, does not i-upport it, but shows why such a pendulum 

 is unsuitable. It is necessary to emphasize this, for it relates to 

 a fundamemal matter in the dynamics of earthquake measure- 

 ment—a matter on w hich the work done of late years in Japan 

 seems to me to be so intimately based that a misunderstanding 

 about it must be fatal to a proper appreciation of that work. 

 And, in point of fact, I did not find that M. Fouque gave an 

 appreciative account of what any of the Japanese observers had 

 done. As (o his mention of Prof. Ewing's seismograph, in 

 particular, I criticized it not so much because it was meagre as 

 because it was incorrect, — so incorrect as to justify the inference 



I that the author was not acquainted with that instrument. 



I The Reviewer. 



I Hertz's Equations. 



' Mr. Watson's criticism, that Hertz's equations are only true 



for places at some distance from the oscillator, is no doubt 

 perfectly valid. [There is, by the way, an insignificant and 

 obvious misprint of A. for p about the middle of his letter.] But 

 this was entirely recognized by Hertz himself; he treated the 

 oscillator as infinitesimal, knowing that it was nothing of the 

 kind when you got near it, and refrained from drawing his 

 diagram-curves into its neighbourhood, for this very reason. 



The fact is surely that, to work out completely the case of 

 electric oscillators in a compound body formed of a couple of 

 spheres joined by a cylinder, would tax the resources of a strong 

 mathematician ; and it is impossible that the vibration can be^ 

 in any sense, a pure one ; all manner of sub-vibrations must bs 

 superposed upon the main. 



From the physical point of view, some general notion of what 

 was happening at a distance of a wave-length or more from the 

 oscillator was desirable, and this Hertz satisfactorily obtained. 



But, to work out what is happening in the immediate neigh- 

 bourhood of a dumb-bell oscillator must be left, I imagine, to 

 the time when some pure mathematician may devote his atten- 

 tion to this particular shape of conductor, if the case appears to 

 him of sufficient interest. At present I see no special reason 

 why it should be so regarded, but of that Mr. Watson is a better 

 judge. I hope he may see fit to attack the problem. 



Grasmere, April 13. Oliver J. Lodge. 



THE COMPRESSIBILITY OF HYDROGEN. 



A S stated in the obituary notice that appeared in 

 ■^^^ Nature (vol. .\xxviii. p. 59S) at the time of the 

 melancholy accident which caused his death, Wroblewski 

 was engaged in an investigation of the behaviour of 

 hydrogen on compression. The results of this investi- 

 gation, as far as it had then advanced, have now been 

 made public {Monatsh.fiir Chcm., 1888, p. 1067 ct seq.). 

 They are of a most important and interestmg nature, and 

 form a fitting memorial of the patience and skill of the 

 observer, who most unhappily was not spared to bring 

 this, the last and most complete of a long series of 

 similar investigations, to a close. 



Hydrogen has long occupied an exceptional and iso- 

 lated position among gases. This is due to the fact that, 

 as Regnault first pointed out, hydrogen forms the sole 

 e.xception to the law that the product of the pressure into 

 the volume, pv, of any gas decreases with increasing 

 pressure, — the exact converse being true in the case of 

 hydrogen, this product showing a regular increase. It is 

 true that, as since shown by Amagat and others, this 

 behaviour of hydrogen becomes general for all gases 

 when the pressure is increased beyond a certain limit, 

 but before reaching this limit the product pv invariably 

 decreases until a minimum is reached for all gases with 

 the exception of hydrogen. P"or hydrogen neither the 

 decrease nor the minimum have yet been observed, the 

 gas as hitherto examined showing an invariable increase 

 of pv with increasing pressure. The natural inference 

 was, however, that the exception was only apparent, and 

 that the minimum above noted would be found to occur 

 also with hydrogen if the gas were examined at lower 

 pressures than those hitherto investigated — that is to say, 

 at pressures below one atmosphere. But a difficulty in 

 the way of this hypothesis arises from the fact that the 

 critical pressures of all gases are found to be below the 

 pressure at which the minimum value for the product of 

 pressure into volume occurs, and therefore on the above 

 reasoning the critical pressure of hydrogen would have 

 to be phenomenally low and considerably beneath one 

 atmosphere. 



To gain a further insight into the relation of volume to 

 pressure in the case of hydrogen, Wroblewski decided to 

 investigate this relation through a wide range of tempera- 

 ture. For this purpose he selected as temperatures suffi- 

 ciently apart, the boiling-point of water, 100" C, the 

 melting-point of ice, o^ C, the boiling-point of liquid 

 ethylene, - io3''-5 C, and the boiling-point of liquid 

 oxygen, — 183° C. The pressures employed varied from 

 one to seventy atmospheres. 



The method of experimenting was exceedingly simple. 

 The gas at a known pressure was forced into a bulb of 

 known capacity having a capillary neck, and kept at one 

 of the above four temperatures. A sufficient length of 

 time was allowed for the gas to attain the fixed tempera- 

 ture ; it was then transferred to a eudiometer, and its 

 volume measured. It is needless to add that every pre- 

 caution was taken both in purifying the gas and in 

 applying the necessary corrections. 



The results with the three first of the above tempera- 

 tures agree with the behaviour of hydrogen already 

 observed, the product of volume into pressure constantly 

 increasing with the pressure. It was found that for the 

 range of pressures under investigation (one to seventy 



