August 30, 1906] 



NA TURE 



443 



sent in use in various countries are not always those 

 ordinarily adopted. In Japan, for example, the pre- 

 sent standard of mass is the " Kwan," prototypes of 

 which were recently standardised at Sevres. 



We can, however, cordially recommend the book, 

 which should prove very useful. J. A. H. 



LETTERS TO THE EDITOR. 

 [The Editor aoes not hold himself responsible for opinions 

 expressed by his correspondents. Neither can he undertake 

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

 manuscripts intended for this or any other part of Nature. 

 No notice is taken of anonymous communications.] 



Thermodynamic Reasoning. 



In the address delivered by Principal Griffiths at York, 

 which is printed in your issue of August 9, I read : " Prof. 

 Armstrong remarks that it is unfair to ' cloak the inquiry 

 by restricting it to thermodynamic reasoning, a favourite 

 manauvre with the mathematically minded.' He adds 

 that such a course may satisfy the physicist but ' is 

 repulsive to the chemist.' The inquiry, ' Why is the 

 application of thermodynamic reasoning repulsive to the 

 chemist?' naturally suggests itself." 



This statement shows a strange misapprehension of my 

 position. I have taken exception to the restriction of the 

 inquiry to thermodynamic reasoning, not in any way to 

 the mere application of thermodynamic reasoning. My 

 objection was to formula worship. I still and shall ever 

 object to it, for it is the bane of progress. As I said at 

 York, physicists too nearly resemble the visitors to London 

 who walk along the Strand and Shaftesbury Avenue and 

 are eontent to look at the theatres from outside ; thev 

 resemble those who admire the British Museum building 

 but have no desire to examine the treasures within it. 



If I did not misunderstand him, Mr. Whetham implied 

 at York that it was enough for him that a certain thermo- 

 dynamic expression was valid : what the condition termed 

 osmotic pressure really is — whether a true pressure or 

 whether, as I suggested, a negative pressure or thirst — 

 mattered not a jot. A certain mathematical thermodynamic 

 picture being painted, no other artist need apply. This 

 does not seem to me to be the attitude a scientific inquirer 

 should adopt. Whether I represent the opinion of chemists 

 matters little : personally I am not willing to remain out- 

 side the Museum : I shall go inside, if possible, trusting 

 that in some faint degree I may be able to appreciate the 

 wonders within it. 



.•\t present, progress is not a little hampered by the fact 

 that chemists and physicists cannot wander through the 

 museums of nature looking eye to eye in complete sym- 

 pathy with one another : surely we are destined to be the 

 closest of friends ; more should be done to cultivate an 

 understanding : a confusion of tongues has arisen which 

 keeps us apart : we must both strive to speak a simpler 

 language. Together 



Henrv E. Armstrong. 



It is the strength and weakness of thermodynamical 

 reasoning that it connects different phenomena without the 

 aid of theories about the mechanism by which the connec- 

 tion is effected. 



In the discussion at York, Prof. Armstrong put forward 

 certain arguments in favour of the view that solution is a 

 chemical phenomenon, and osmotic pressure due to an 

 attraction of the nature of chemical affinity. He used these 

 arguments in an attempt to invalidate van 't Hoff's 

 ther'Tindynamic theory, which shows that, from the observed 

 solubilifv phenomena of volatile substances, it follows that 

 the ideal osmotic pressure of a number of partiel'^s of 

 such substances in a dilute solution must be equivalent 

 to the pressure which the same number of particles would 

 exert as a gas occupying the same space. 



In my reply to Prof. .Armstrong I pointed out that the 



NO. IQ2 2, YOL. 74I 



thermodynamic theory is quite independent of the particular 

 view we may adopt as to the fundamental nature of solu- 

 tion, and the modus operand^ of osmotic pressure. Osmotic 

 pressure may, as van 't tloff himself supposed, be due to 

 the impacts of the dissolved molecules ; it may, as Prof. 

 .•\rmstrong believes, be caused by chemical affinity ; it may 

 be produced by some other undiscovered cause. The 

 thermodynamic reasoning avoids all such hypotheses, and 

 connects directly the experimental facts of the solubility 

 of gases with the osmotic pressure they would exert 

 against a perfect semipermeable membrane in dilute solu- 

 tion. 



I have never suggested that the ultimate nature of solu- 

 tion was a matter of no interest. It is the question of 

 most supreme importance now outstanding in these subjects ; 

 but let us clear the issue before attacking it. We must 

 recognise clearly that the relations indicated by thermo- 

 dynamics and 'confirmed abundantly by experiment are 

 among the established facts to be explained by a theory of 

 the nature of solution. 



It is for this recognition of the true position of the 

 problem that I contend. The thermodynamic reasoning 

 which connects the ideal osmotic pressure with experi- 

 mental phenomena is not in question. That reasoning is 

 confirmed by measurements of actual osmotic pressures 

 and of freezing points. It can only be invalidated by a 

 general attack on thermodynamic theory, such as that 

 which was foreshadowed in Mr. Campbell's recent 

 reconnaisance-in-force. I do not think any such attack 

 has much chance of success. Osmotic phenomena seem 

 to me to be entrenched in the strongest part of the vast 

 lines occupied by the science of thermodynamics. 



Cannot Prof. Armstrong agree to accept the thermo- 

 dynamic reasoning as confirmed by experiment, and pass 

 on to the further problem? Personally, I think that the 

 evidence at present available is on the whole in favour 

 of the chemical theory of solution and osmotic pressure 

 —the theory which Prof. Armstrong supports; but there is 

 work to be done before such a conclusion can be taken as 

 established. May we not agree that it is better both for 

 physicists and chemists to do such work than to waste 

 their energies in attacking with inadequate artillery the 

 well-fortified citadel of thermodynamics? 



W. C. D. Wheth.^m. 



High Borrans, Westmorland, August 21. 



The Iron Arc. 



While carrying on some experiments with the electric 

 arc between iron electrodes, one of my students, Mr. H. D. 

 Arnold, noticed that there was a certain critical P.D. at 

 which an abrupt change took place in the conditions of 

 the arc. Subsequent investigation has shown that the effect 

 is closely analogous to the " hissing point" of the carbon 

 arc. How close the analogy is may be seen from the 

 following remarks. If the iron arc is started with a large 

 external resistance and maintained at such a length that 

 the current is well below one ampere, it burns with little 

 or no sound, and its appearance in the neighbourhood of 

 the anode is very diffuse and ill-defined. As the external 

 resistance is gradually decreased, the P.D. falls and the 

 current rises until a certain critical value, depending on 

 the length of arc and size of electrodes, is reached. At 

 this point a very small decrease in external resistance 

 suffices to cause a sudden increase in current and drop in 

 P.D., precisely as with the carbon arc. At the same time 

 the arc contracts, a bright spot appears on the anode, and 

 a characteristic hissing sound begins. Further increase of 

 current is accompanied by a continued decrease in P.D. 

 The hissing stage, in fact,' begins at quite a different point 

 on the P.D. -current diagram from that in the case of the 

 carbon arc. If the experiment is carried out in the reverse 

 order, starting with a large current, the discontinuity is 

 encountered again, but not until the current has been 

 diminished beyond the value that it had at the beginning 

 of the hissing stage. Indeed, with arcs of 6 mm. and 

 more, the current on the hissing stage can with care be 

 decreased until it is smaller than its previous largest value 

 on the quiet stage. Thus there are two possible values 

 of P.D. for the" same current and length of arc, one 

 corresponding to the quiet, the other to the hissing stage. 



