March 3, 1892] 



NATURE 



415 



L'effet Thomson a ete mis en evidence par I'experience. M. 

 Tait croit qu'il en est de meme de la difference de potentiel 

 vraie. 



Ou la phrase que j'ai citee plus haut n'a aucun sens, ou elle 

 signifie qu'il me blame d'avoir dit le contraire. 



Or cette maniere de voir ne soutient pas un instant d'examen. 

 Nous n'avons aucun moyen de mesurer la difference de potentiel 

 vraie. 



Les melhodes electrostatiques ne nous font connaitre que la 

 difference de potentiel apparente ; les methodes electrodyna- 

 miques ne nous font connaitre que la somme des forces electro- 

 motrices vraies dans un circuit y^rW, 



Enfin les methodes indirectes, fondees sur I'ecoulement ou 

 sur les phenomenes electrocapillaires, ne sont pas applicables 

 dans le cas qui ncus occupe. H. Poincare. 



The Theory of Solutions. 



It seems that, unfortunately, the period of misconceptions, 

 whose victim the theory of solutions is, has not yet ended. For, 

 after an explanation from my side of the theory of solutions as 

 I understand it, Mr. J. W. Rodger, my critic, asserts (Nature, 

 p. 342) that •' it cannot be admitted that a number of exact rela- 

 tionships constitutes a theory." From his further remarks, it 

 must be concluded that he designates by the name theory what 

 I would name a hypothesis, and that, according to him, van 't 

 Hoff's application of the " gaseous laws" to solutions involves 

 the hypothesis that there exists no interaction between the 

 solvent and the dissolved substance. 



It was therefore in vain that I stated in my letter, in italics, 

 that many properties of the solutions, according to the new 

 theory, ' ' can he treated entirely independently of the question of a 

 possible interaction betiueen the parts of the dissolved substance 

 and the solvent " ; it was in vain that I pointed out that all the 

 laws concerning these properties are solely consequences of the 

 one law relating to the volume energy to be gained by making 

 up a solution. This law, whose expression is pv = RT, in its 

 various applications to solidification, vaporization, osmosis, &c., 

 of solutions, is the issue of a great many special laws, the 

 whole of which I name the new theory of solutions. Such a 

 complex of laws, grouped around and derived from a main law, 

 is what I call a theory ; and if the theory, as in the present case, 

 is everywhere in accordance with experience, the main law is to 

 be regarded as correct. There is nothing of hypothetical nature 

 in this theory, for, if once the main law, pv = RT, is given (by 

 osmotic experiments or otherwise), all the special laws are 

 merely thermodynamical consequences of it. And, I repeat, 

 the main law involves no hypothetical assumption upon the 

 mutual role of solvent and dissolved substance, but is solely the 

 condensed expression of a great number of experimental facts 



Mr. Rodger asks why I did not state clearly in my book that, 

 in ray opinion, interactions between solvent and dissolved sub- 

 stance were possible. I can only reply that on suitable occasions 

 I have done so. Besides the sentences quoted by Mr. Rodger 

 himself, I have devoted (pp. 251, 252) half a page to the evi- 

 dence that considerable interaciions take place in salt solutions 

 on dilution. But as the existence of such interactions, as I 

 have shown, is of no consequence in the statement of the general 

 laws, I have treated them as secondary, however interesting 

 they may be as experimental facts, and I am more than ever 

 persuaded by this discussion that I was right in doing so. For 

 I have not written my book for readers prepossessed by some 

 non existing chemical theory of solutions, but for such as wish 

 plainly to learn what is known about solutions. 



Similar remarks are to be made as to the definition of solu- 

 tions as mixtures. Even in the case of interactions, if, e.g., 

 hydrates are formed in a solution, the solution is finally a mix- 

 ture of the hydrates and the remaining solvent. For the con- 

 trary assumption — that the whole of the solvent is combined with 

 the dissolved substance, that, e.g., in a somewhat diluted solution 

 of common salt, there exist compounds, as NaCl + 1000 HoO — 

 is in such a degree at variance with all known facts that 1 did 

 not think it worth while to dij-cuss such an idea. 



Lastly, Mr. Rodger terms the application of the formula of 

 van der Waals to solutions a-; in general " highly questionable " 

 and as " meaningless," if it is admitted that "something of the 

 nature of a chemical reaction " between solvent and dissolved 

 substance may occur. Mr. Rodger may convince himself from 

 my book that this application is limited to cases in which I do 

 not suppose the occurrence of chemical reactions. The reasons 



NO. II 66, VOL. 45] 



of his doubts as to the validity of this application I cannot 

 remove, because he has not stated his reasons. But it may be 

 permitted to me to feel some doubts as to the validity of his 

 reasons. For no other than van der Waals himself has taken 

 up this very question, and has discussed (of course much more 

 fully than I was able to do) the application of his formula to 

 solutions, including also the case of interactions between the 

 substances. His papers on this subject are inserted in the 

 Zeitschrift fiir physikalische Chemie, v. p. 133, and viii. p. 188 ; 

 and also in the Archives Neerlandaises of 1889 and 1891. 

 Leipzig, February 16. W. Ostwald. 



A Lecture Experiment on Sound. 



The following experiment may be of interest to your readers. 



A piece of glass tubing is drawn out to a fairly fine point, P, 

 attached by string crosswise to a short lath of wood, \v, con- 

 nected by india-rubber tube to water-tap, and a jet of water 

 directed on to a tambourine, T. 



A tuning fork held in one hand is made to touch the lath held 

 in the other while vibrating, and the whole moved nearer to or 

 further from the tambourine. 



At a certain distance the note of the fork will be produced on 

 the tambourine (this of course is not a new experiment). 

 While this was going on, the lath, jet, and fork were slowly moved 

 towards the tambourine, and I was able to sound the octave 

 below. 



This showed that at a certain point the vibrations of the fork 

 were not individually capable of separating the fine stream into 

 drops, but that two complete vibrations did so ; thus half as 

 many drops per second were set free as there were vibrations 

 from the fork. 



The fork gave C =512; the note on the tambourine was 

 C = 256. 



Probably the drops at that stage were of a dumb-bell shape — 

 since at a greater distance the actual note of the fork was 

 produced on the tambourine. Reginald G. Durrant. 



The College, Marlborough, February 13. 



The Formation and Erosion of Beaches, &c. 



As you have more than once permitted me to discuss the 

 problem of sea-waves in your columns, I venture to point out 

 that in your interesting article on Signor Cornaglia's work on 

 sea beaches (p. 362), in your summary of the causes which affect 

 beaches, sand-banks, &c., you have omitted the very important 

 one of wind-raised surface currents. Sea-waves, tidal-currents, 

 and river-currents can be observed, and their effects recorded ; 

 but it is the occasional, irregular, and sometimes powerful wind- 

 raised current, prevalent during storms, which performs such 

 erratic feats, and deludes the unwary observer. For instance, 

 a beach may resist the sea for years, yet in a few hours it may 

 be stripped bare to the solid rock. Shells may be covering the 

 bottom a mile off shore, undisturbed by on-shore gales ; a 

 storm, with wind and waves apparently much the same as usual, 

 may sweep them all on shore. One beach will be invariably 

 kept clear of shells which will be found off shore, while 



