10 EXT'ERDIENl-S Wnil IONIZED AIIJ. 



color i)ri)(luciiig, though opaques are rare. It is a striking feature of this experi- 

 ment, that tlie air of a room \vhen tested with the color tube remains permanently 

 fouled for lioui.s afterwards. In all these instances au unstable sulphur compound 

 is apparently preserved from oxidation, and this nucleus is permanently present in 

 coal gas. 



11. — It is well to note that such volatile chemicals as camphor, camplior mono- 

 bromide, naphthalene, lienzoic acid, phenol, toluol, amylbenzoate, diphenylamine. 

 etc., like sal ammoniac, produce no color effect unless they are charred. Some of 

 these are electrolytes. With ammonic chloride it is even possible to clear the blue 

 Held due to phosphorus, by passing its emanation with a definite amount of the 

 smoke of the former into the color tube. The phosphorus nuclei seem thus to be 

 absorbed by the particles of sal ammoniac and the condensation effect of the 

 foi-mer is desti'oyed. This result seems to me to be of considerable importance. 



If coal gas is passed through tlie phosphorus tube, figure 9b, the field is all 

 but cleared aftei- an initial puff' of color. There is, however, a residual faint tint 

 due to the coal gas alone. Tlie phosphorus at once becomes active again when air 

 is blown through the tube. 



12. Electrical ionizers. — The prime conductor of a Holtz machine placed near 

 the mouth of the color tube is without effect. When a needle is added, however, 

 the coloration becomes intense. ITeiice the brush or glow is a powerful dust pro- 

 ducer. Air })assed through the cylindiical shell of a tubular condenser is not active 

 without sparks. If the air is dust laden it is not appreciably affected by the 

 condenser. 



A gauze condenser sui'mounting the color tube C, as in figure 12, shows intense 

 action corresponding to the continual leakage. If enclosed by a close-fittini; glass 

 jacket, D Z>, figure 1 3, its effectiveness is nearly destroyed, probably because the leak- 

 age runs down the glass or the lines of force terminate in it. In these experiments 

 a spaik at the machine clears the field completely, so that the color surges with the 

 potential. If the upper condenser plate is a spiral brush of fine wire frauze, ,% 

 figure 14, the color effect continues for a distance of even 20 centimeters between 

 the plate-s, the machine now giving one-inch spai-ks. Similarly \ inch sparks on the 

 machine give like color results for ^ inch distances between the plates. 



Moie effective than a narrow jacket is a wide bell jaj-, as shown in fi^ui-e 15. 

 Here \ inch sparks on the machine darken the field when the condenser plates are 

 20 or even 30 centimeters apart. Reversal of polarity showed no definite results. 

 Darkness now continues during sparking at the machine, and it takes about 30 

 seconds to empty the vessel after the electric machine is switched out. Here and 

 elsewhere a maximum darkening effect seems to cori'espond to a definite spark 

 length. Very high jiotentials prol)ably develop brushes along the electric conduc- 

 tors, as the ])resence of an electric wind not playing into the tube C, clears the field. 



^•^- — The electric duster is further increased in efficiency by the spiral brush 

 >S and ladial condenser arrangement of figure 16. In the dark the influx into the 

 color tube C pa.sses through the radial glow and becomes a tremendously strong 

 dust producer, even with small sparks on the machine. 



