M 



\Y. I904..J 



KNOWLEDGK & SCIENTIFIC NEWS. 



79 



you do this the vacuum tube will go out and the discharge 

 will not be re-established at the spark gap till you remove 

 the radium salt. 



V\g. 4- 



The coil must not give too strong a discharge, and the 

 discharge must be very steady. If this is secured, the 

 experiment can be brought off with great certainty every 

 time. 



{To be coniinufd.) 



Modern Views of 

 CKemistry. 



By H. J. H. Fenton F.R.S. 



A FEW further illustrations may be given of the simple 

 explanations which the ionic-dissociation hypothesis 

 affords of the properties and reactions of salts in solution. 

 Smce salts are highly ionised when dissolved in even a 

 moderate quantity of water, the properties of the solution 

 represent the joint or added properties of the ions into 

 which the salt has split up. The colour of the solution, 

 for example, is that of its ions ; solutions of most common 

 cupric salts are blue and nickel salts green. This is 

 because the acid-radicles (sulphate, nitrate, &c.) happen 

 to give colourless ions and the colours observed are 

 therefore due to the metallic ions. Most permanganates 

 are pink and manganates green in solution, because the 

 metallic ions (potassium, sodium, &c.) happen to be 

 colourless and the colours here are due to the acidic ions. 

 It is interesting to observe in this connection that some 

 ions may be correctly represented by the same chemical 

 symbol and yet show different colours and other proper- 

 ties in solut'on. Both the manganate ion and the per- 

 manganate ion are represented by the symbol ^MnOj, 

 yet one is green and the other pink. The copper 

 ion again is blue when in the cupric state, but colour- 

 less in the cuprous state. This is explained by saying 

 that the electric charge associated with the ion is 

 different in the different states. A well known and 

 simple experiment in illustration of the above views may 

 be made as follows : Dissolve some dry cupric chloride, 

 which is brownish yellow, in a very little water ; the 

 solution appears green. Dilute it, and it becomes blue ; 

 add a strong solution of hydrochloric acid or sodium 

 chloride and it turns green again. Repeat the latter part 

 of the experiment, using mercuric chloride instead of sodium 

 chloride, and the solution remains blue. The "ionic" 

 explanation is that the very strong solution first made 



contains some molecules of cupric chloride (impure 

 yellow) mixed with copper ions (blue) and tliis mixture 

 gives to the eye the appearance of green. On diluting, 

 the cupric chloride molecules are further ionised, giving 

 therefore less yelk)w and more blue. If, however, a 

 strong solution of a metallic cliloride is added, its chlorine 

 ions, being in great concentration, prevent the further 

 ionisation of the cupric chloride, according to well known 

 principles which will be discussed later. It happens, 

 however, that mercuric chloride is an exceptional salt in 

 that it is only very slightly ioniseil when it is dissolved in 

 water; there are scarcely any free ciilorine ions in its 

 solution therefore, and it can have little influence in 

 checking or reversing the ionisation of the cupric chloride. 



The colour-changes of the indicators which are used 

 in analysis, such as litmus, may be explained in a similar 

 way. We may regard these indicators as behaving like 

 very weak acids and the colours they show in acid solu- 

 tions, where they are very little, if at all, ionised, is the 

 colour of the compound or molecule — red in the case of 

 litmus. But now on adding an alkali a salt is formed, 

 and this, like nearly all salts, is highly ionised in solution, 

 so lliat we now see the colour of the acidic ion — blue in 

 the case of litmus. The colour-changes of other well- 

 known indicators can be similarly explained ; in phenol 

 phthalein the molecule is colourless, tlie acidic ion pink, 

 wOiereas in the case of methyl-orange the molecule is 

 pink and the acidic ion yellow. 



rrhis very simple and attractive cxphuiation of the 

 colour-changes in indicators has, it must he confessed, 

 received rather a severe " shaking " owing to certain 

 recent observations, and it is probable that the effects 

 depend rather upon changes of constitution in the indi- 

 cator.! 



Not only th^ colour but the reactions of a solution of 

 a salt are considered to be due to its ions; a solution of 

 ferrous, or ferric chloride, for example, gi\es a precipi- 

 tate with alkalis due to the iron ion and a precipitate 

 with silver nitrate due to the chlorine ion. Potassium 

 ferrocyanide, however, gives no precipitate with alkalis, 

 although it contains iron, and chloral gives no precipitate 

 with silver nitrate although it contains chlorine. The 

 potas-iuni ferrocyanide contains its iron associated with 

 cyanogen as a complex group, and when dissolved gives 

 potassium ions and ferrocyanide (l'eCr,N(,) ions; none of 

 the iron, as such being present in the ionic state. Cliloral 

 again gives a solution which contains no chlorine ions ; 

 the chlorine is combined with the other elements as an 

 undissociated molecule. 



Mercuric cyanide has long been known as abnormal in 

 its behaviour, since it answers scarcely any of the usual 

 tests either for mercury or for a cyanide. It can he shown 

 in \arious ways, however, that the salt is practically dis- 

 solved unchanged ; its solution contains neither mercury 

 ions nor cyanide ions. The poisonous character both of 

 mercury salts and of cyanides is assumed to be due to 

 their ions ; therefore we should expect mercuric cyanide 

 to be non-poisonous. This is stated to be the case, 

 although it does not appear that any ardent supporter of 

 the " ionic " theory has had the strength of mind to try 

 its effects upon himself. 



The most "chemically active" substances then in 

 solution are those which are most ionised. It does not 

 follow, how^ever, that all chemical changes which may 

 take place in solution are necessarily ionic changes. It 

 has been shown, for example, that certain salts and acids 

 undergo immediate double decomposition when dissolved 

 in solvents in which no ionisation occurs. 



The action of a strong acid upon the salt of a weak 

 acid was formerly looked upon, as indicated above, as 



