114 



SCIENCE. 



[Vol. XXI. No. 526 



(each point similarly situated to every other point) may be ar- 

 ranged is sixty-six, or just the number of the well-defined ele- 

 ments. So that imagination may picture Spencer's homogeneous 

 cloud of atoms splitting up into these different ' ' space net " ar- 

 rangements, each kind of net being a different element.] 



3. Solution. The chief opponent of the disassociation theory 

 of solution is Professor Pickering; and his chief argument against 

 it (for, of course, the disassociation theory allows the formation 

 of hydrates as well as Professor Pickering's own hydrate theory 

 does) is the fact that while disassociation almost always takes 

 place with absorption of heat, solution generally emits it. This 

 anomaly can be explained very satisfactorily by the electrostatic 

 theory of cohesion. For whether a substance is a solid (or fluid) 

 or a gas depends on whether the fraction 

 T ' cohesive force of atoms. 



repulsive force due to kinetic energy ot atoms + at- 

 traction of atoms for other atoms 



is greater or less than unity. We can thus turn a substance into 

 a gas by either decreasing the numerator or increasing the de- 

 nominator. The numerator we cannot change. The first term 

 of the denominator we can increase by heating the substance, the 

 second term by placing the substance in contact with a solvent. 



In the last case the atoms of the solid part company with each 

 other. But their cohesive force is not lost; it is simply added to 

 that of the solvent, as shown by the increase of surface tension 

 and of boiling point of a solution over that of the solvent. Since 

 the solvent takes* up the stress there is no necessary evolution or 

 absorption of heat. A mechanical simile will make my meaning 

 clear: Suppose a spiral spring, A, fixed on a board, C, which 

 when compressed gives out heat from some reversible cause, so 

 that it will absorb the same amount of heat in expanding. This 

 is similar to the behavior of a gas — when compressed it gives 

 out heat, when it expands again it absorbs heat. 



But now suppose a second spring, B, placed beneath the board, 

 C, similar in every respect to the first spring, and its axis a pro- 

 longation of that of A. Suppose an iron rod fastened to the bot- 

 tom of C, extending up the centre of both springs, the rod being 

 somewhat longer than one of the extended springs, and having a 

 hook on the end of it. 



done to plate the dissolved substance out, and the electromotive 

 force necessary to do this, since the amperes are constant for all 

 equivalents, must depend on the rate at which the surface ten- 

 sion varies per withdrawal of unit weight of the electrolysed 

 substance, allowing also for any heating or cooling during the 

 electroplating. 



3. Compression of gases. The ordinary formula for the com- 

 pression of gases is that of Van der Waals, i.e. : — 



I. 



(i' + f.) (v-i>)=Rr. 



If the electrostatic theory of cohesion is correct, the equation 

 should read 



II. 



(p+^ (^v-b'J ^ar. 



for reasons evident to those who have read the previous note 

 {Science, Aug. 22, 1893). 



This is no longer a cubic, and it is pretty certain that the equa- 

 tion for the compression of gases should be one on account of the 

 shape of the pressure-volume curves of carbonic acid gas. But 

 we can transform the above equation, II., into a cubic by putting 

 a, no longer as a constant but equal to a constant multiplied by 

 .yf. The equation thee reads : — 



III. 



p+ L^) (t;-6)=ijr. 



in which c is the same for all gases. The experimental data agree 

 with this modified equation, as shown by table I. 



Table I. 



^S- 



In Fig. 1 both springs are extended. In Fig. 3, the spring^ is 

 compressed, heat being given out. 



If it is now allowed to expand, the same amount of heat will 

 be absorbed. This latter represents the turning of a solid into a 

 gas by heating it. 



But suppose, being compressed, the iron rod is hooked over the 

 top of it. Then when it is let go it will expand and assume the 

 position of Fig. 3. But no heat will be generated in the system, 

 for it is evident that B will give out just as much as A absorbs. 

 If the amount of heat given off by unit contraction of A were 

 greater than that given off by B, the resultant effect would be 

 a cooling of the system. If it were less, the resultant would be 

 a heating. So we see, that while the expansion of A by itself 

 would always absorb heat; when it is joined to B, the resultant 

 effect depends on B. 



Now, this is a very fair simile of what goes on,]when a solid is 

 dissolved in a solvent. The solid loses its stress, which is taken 

 up by the solvent, the result being an increase of cohesion be- 

 tween the molecules of the solvent, producing as a natural con- 

 sequence increase of surface tension, lowering of th e freezing 

 point, and raising of the boiling point. 



If the added electrostatic strain produces a greater amount of 

 heat in the solvent than the loss of strain in the solid would ab- 

 sorb heat, the resultant would be a heating of the whole solution. 

 Since, when a dissolved substance is plated out by electrolysis, 

 the result resembles the cutting of the iron rod, D, in Fig. 8, 

 there is an absorption of energy (5r cooling, so that work must be 



Substance. 

 Dyethylamine 

 Ethyl. Acet. 

 Ether 

 Benzine 

 Ethyl. Form. 

 Chloroform 

 Acetone 

 Methyl. Acet. 

 Alcohol 

 Ethyl. Chlor. 

 CSs 

 SO2 

 NO2 

 as closely as can be expected 



This table shows that a varies as (volume) ^. Two substances 

 do not agree with this theory — benzine and NOj. This is owing 

 to the fact that the data are given wrongly in the table from 

 which this is copied (i.e., that in Ostwald's "Outlines of General 

 Chemistry"). This is seen^by the following facts. From the 



