414 



NA JURE 



[March 6, 1902 



generate other ions by collisions with the molecules of the gas 

 when they move sufficiently rapidly. 



(/') For any gas the negative ions which are generated by 

 collisions are the same (having the same properties over wide 

 ranges of force and pressure) as those which were generated by 

 the rays. 



(c) If o is the number of negative ions generated in a gas by 

 one of these negative ions moving through one centimetre, then 



where / is the pressure of the gas and X the 



'^(>> 



electric force acting on the ion. 



(rf) The free paths of the negative ions are long, compared 

 with the free paths of the molecules, so that their linear dimen- 

 sions must be smaller than those of the molecules. 



From a comparison of the properties of positive and negative 

 ions, it can be seen that the mass of the negative ion must be 

 small compared with that of the positive ion. 



The values of a were determined for large ranges of pressure 

 and electric force for air, carbonic acid and hydrogen, and the 

 functions / have been represented graphically by three curves. 

 Let /,, /, and /, denote the functions / found for air, car- 

 bonic acid and hydrogen respectively as determined by the 

 experiments with Rontgen rays. 



The results which I have to add to these were obtained by 

 finding the conductivities of gases between parallel plates when 

 one of the plates is illuminated by ultra-violet light. 



The experiments have led to the following conclusions : — 



(a') The negative ions set free from a zinc plate when ultra- 

 violet light falls on it generate other ions by collisions with 

 molecules of air, carbonic acid or hydrogen. 



(b') The negative ions thus generated by collision in the gases 

 have the same properties (over large ranges of pressure and 

 electric force) as the ions generated by the light from the zinc. 



[Hence the.se four kinds of ions are identical, viz., the ions 

 given off from the zinc plate and the negative ions which they 

 produce in air, carbonic acid or hydrogen. These negative 

 ions may be denoted by the letter Z.] 



((■') If a' is the number of ions which one of the 7. ions pro- 

 duces per centimetre by collisions with molecules, then a is 

 connected with the electric force and the pressure by an equation 



of the form a' — p/'{'—\ The three functions //, f,^, f.^ as 



determined in this manner for air, carbonic acid and hydrogen 

 are equal respectively to the corresponding functions /j, yo../:; 

 as determined by the experiments with Rontgen rays. The 

 equality extends over the whole ranges of pressures and force 

 which have been examined. 



Consequently the negative ions generated by Rontgen rays 

 in a gas are precisely the same as the ions set free from a zinc 

 plate by ultra-violet light. 



If it be questioned that the identities /,=// ;/,=/,' ;/,=/,' 

 are sufficient to justify this conclusion we may proceed to establish 

 the proposition in the following manner : — 



The charges on negative ions produced by Rontgen rays in any 

 of the gases under consideration have been shown to be equal to the 

 charge on a negative ion given off from a zinc plate by ultra-violet 

 light (J. S. Townsend, P/iil. Trans. 1899 and igoo). 



For simplicity, one of the gases may be considered, air, tor 

 example. Let m be the mass and e the charge on a negative 

 ion R produced in air by Rontgen rays, and let //;' and e' be 

 similar quantities for an ion Z produced by the aid of ultra-violet 

 light. 



Since the maximum values of/, and/,' as determined by the 



larger values of — are equal the two kinds of ions R and Z 



must produce the same number of collisions per centimetre 

 so that they have the same free paths. For any force \, the 

 kinetic energy that the R and Z ions acquire along their free paths 

 must be equal since their charges are equal. At the end of 



a path of length .v the value of "'— or "'^ is equal to Xc.v. 



Hence we have the equation mv-=m'v"-, v and v' being the 

 velocities of the ions R and Z before collision. 



A second independent equation is obtained when we consider 

 the identity/, =/,'. 



The chance of producing new ions by collision is not de- 

 termined by the energy of the colliding ion. If this were the 

 case, the positive ions would produce others by collision under 

 an electric force X if their mean paths became equal to the 



NO. 1688. VOL. 65] 



mean paths of the negative ions when they generate others 

 under the action of the force X. It is easy to show that the 

 positive ions do not acquire the property of producing others by 

 collision even when their free paths are much longer than those 

 of the negative ions when they are giving others by collisions. 

 The negative ions therefore possess this ionising property in 

 virtue of the large velocities they acquire along their free paths. 

 It is therefore evident that the function / involves the mass and 

 velocity of the colliding ion in .some form which is not reducible 

 to the product in x v'-. The equality of/, and /,' for the same 

 values of e, X and / supplies us with an equation between 

 »i, 7!, ?ii\ v\ of the form <f(;«, v) = <p[m', v'). Combining this 

 equation with the equation mt/' = m'v'-, we see that m = vi 

 and V — v' , ■ ■, 



Hence the masses of the two ions R and Z are the same as 

 well as their free paths and charges. We thus see that it is 

 possible, by various methods, to detach negatively charged 

 particles from the molecules of gases which are small compared 

 with the molecules, and that the particles which are detached 

 are the same from whatever gas they are removed. 



JOll.N S. TOW.NSEND. 



New College, Oxford, February 2S. 



The Recent Fall of Red Dust. 



Some observations made last autumn in Cornwall may throw 

 light on the fall of dust in South Wales. On September 2, 

 during gusty weather with squalls from the] E.N.E., I watched 

 from my window at Carbis Bay (270 feet .above the sea) puffs 

 and swirls of dust rising from the desert-like fiat at the mouth of 

 the Red River. The dust-cloud rose above the top of Godrevy 

 Towans (230 feet), nearly blotted out Godrevy Lighthouse and 

 then spread in a well-defined belt across St. Ives Bay for more 

 than three miles to near St. Ives Head, which it must have passed, 

 though this part of the track was invisible from my point of 

 view. A fortnight earlier a similar observation had been made 

 under identical conditions by Mrs. Reid. On neither occasion 

 did the wind reach the force of a gale, it was merely a strong, 

 dry east wind. 



The red mud which gives its name to the Red River is mainly 

 slime produced by the crushing of the tin-ore in the stream-tin 

 works. This mud spreads far and wide over the alluvial flats 

 and along the sandy shore ; when it dries it forms an almost 

 impalpable dust. Much of this dust is mixed with the Cornish 

 sand-dunes, and drifts to and fro with the shelly sand, which 

 forms the main part of those dunes. If the dust-falls in South 

 Wales are of Cornish origin, the material will probably contain 

 a good deal of finely powdered schorl, which mineral occurs 

 abundantly in the tin-ore. Clement Reid. 



The Validity of the lonisation Theory. 



The number of Nature which appeared on January 30 

 contains an abstract of a paper by L. Kahlenberg entitled 

 " The Theory of Electrolytic Dissociation as viewed in the 

 Light of Facts recently Ascertained." In the paper referred to 

 is a kind of summary of observations which have been made on 

 non-aqueous solutions, from the consideration of which the 

 author draws the conclusion that the electrolytic dissociation 

 theory is untenable in the case of non-aqueous solutions. 



In view, however, of the generally accepted opinion that this 

 theory is in good accordance with experimental observations on 

 aqueous solutions, Kahlenberg has been led to investigate such 

 solutions more closely. As the result of a large number of 

 boiling-point, freezing-point and conductivity determinations, 

 the conclusion is drawn that " the difficulties which the theory 

 of electrolytic dissociation encounters in explaining the pheno- 

 mena in aqueous solutions are really insurmountable.'" 



One of the chief reasons for this inference appears to be that 

 the series of molecular weight values calculated from the 

 cryoscopic and ebullioscopic measurements at different concen- 

 trations of the 'solutions are irregular. As an example, it is 

 found that in the case of solutions of magnesium sulphate, the 

 calculated molecular weight, which even in the most difute 

 solution is greater than the theoretical value, increases at first 

 with the concentration, passes through a maximum and then 

 decreases, attaining a value which would correspond to electro- 

 lytic dissociation only in the most concentrated .solutions. The 

 author does not state in what manner these "molecular 



