Fkbkuaky 9, 1900.] 



SCIENCE. 



205 



One way of escaping between the horns 

 of the dilemma is to suppose the ionization 

 to exist even when not directly evidenced. 

 Thus most of the dust with which I have 

 worked is derived by oxidation, or by heat 

 implying chemical action or dissociation, 

 or from highly potentialized matter, etc. 

 Phosphoric dust is quite inactive at low 

 temperatures and in a current of air free 

 coal gas. The same is true of molten sul- 

 phur. Concentrated sulphuric acid, how- 

 ever, shows increased activity when a cur- 

 rent of coal gas is passed through it into 

 the color tube ; but the action here is prob- 

 ably a destruction of the coarse dust part- 

 icles in air, the effect being similar to that 

 referred to in the case of flames. Facing 

 the question squarely it seems extremely 

 difiScult to account for a specific velocity in 

 the non-ionized dust particle. Being al- 

 ready stupefied in o'bserving that a much 

 larger volume of air of initially constant 

 dust content must be discharged through a 

 wide tube than through a narrow tube of 

 the same material, and within certain limits 

 of diameter, instanced by the above table, 

 one may well be daunted in confronting the 

 case of a specific velocity in an inert dust 

 particle. 



The need of a direct decision is, there- 

 fore, urgent ; is this phosphoric dust ionized 

 or not ; or better, is it generated in an ion- 

 ized region ? Using an electroscope, which 

 in the dry room in which I worked retained 

 its charge indefinitely, I aspirated a current 

 of air across the charged terminal without 

 appreciable result. I then blew air which 

 was passed over phosphorus across the ter- 

 minal, and found to my surprise that even 

 with a small current the charge was dis- 

 sipated more than one-half in the first min- 

 ute. It made no difference whether a posi- 

 tive or a negative charge was on the elec- 

 troscope. Hence, phosphorus, when emit- 

 ting condensation-producing dust seems 

 also to emit some form of obscure radiation ; 



for the dust evolution * occurs in a strongly 

 ionized region, or the dust is itself ionized. 

 Eelative to the experiments of the above 

 table, therefore, fresh means are at hand 

 for computing the velocity of the phosphoric 

 dust particle, and the electricity carried 

 per gram, in a way similar to the famous 

 researches made at the Cavendish Lab- 

 oratory, independently of the preceding 

 results. This suggests one method of 

 standardizing the colors of my color tube, 

 absolutely in terms of the number of dust 

 particles per cubic centim. producing the 

 color effect. 



With the question thus happily answered 

 in the case of phosphorus, I next examined 

 air passed over concentrated sulphuric acid, 

 but found it without effect even on the 

 electrometer. Though the acid is a weaker 

 dust producer than phosphorus, the present 

 electrical result seems out of proportion 

 to the data of the color tube. The de- 

 cision made is therefore partial, but I have 

 not advanced beyond it. Other dust pro- 

 ducers might be instanced in the same 

 category, and it is for this reason that I 

 have retained the antiquated designation 

 ' dust ' in this paper. 



I may add that in the experiments which 

 I have under way, I have been dealing 

 somewhat extensively with questions of the 

 above nature, using dust either ionized or 

 not. My methods, however, are all re- 

 stricted to an application of the steam jet, 



* I shall show elsewhere, and have since verified 

 experimentally, that if air energized by phosphoric 

 dust be maintained between the plates of an air con- 

 denser, the difference of potential, ^, vanishes accord- 

 ing to the equation 



tp ^(pQ e— "^■<'* 



in the lapse of time t, where e is the charge per par- 

 ticle. For different thicknesses, d, of air, the decay 

 of energized dust particles in successive layers of air 

 is very similar to the absorption of light in successive 

 thicknesses of medium, viz., n^n^lO — <^d, In my 

 experiments I found a = .25, and thus the number of 

 particles is reduced to 1/10 in a layer of air 4 centi- 

 meters thick. 



