Sept. 19, 1878] 



NATURE 



549 



that the difference in the length of the spark does not 

 bear any precise ratio either to the specific gravity of the 

 gas or its viscosity in reference to mechanical impulse ; 

 they propose to ascribe it, at all events provisionally, to a 

 difference of electric viscosity. In order to make experi- 

 ments on the dge ischarin different gases, they placed the 



Fig. 13. — Arc between Charcoal Points. 



discharger already described (Fig. 4), under a bell glass 

 G G' (Fig. 14) open at the top and covered with a glass 

 plate P. The glass plate had two screw-clamps which 

 were connected at its under surface with wires led from the 

 screw-clamps c and cf of the discharger. In connection 

 Avith these, on the outside surface, were two other screw- 



FiG. i<.—Apparatus for Measuring Length of Spark in Different Gases or 

 at Different Pressures. 



clamps cc' with which the terminals of the battery tt' 

 were connected. Through a stuffing box in the glass 

 cover a steel rod R passed ; this, below the glass, carried 

 a crutch M, with two ebonite pins, which drop into cor- 

 responding holes made to receive them in the micro- 

 meter wheel A ; the rod had on the top, outside the jar, 



a cross handle h for turning it. The distance of the 

 terminals was easily regulated by means of this rod, as 

 the micrometer could be read through the bell glass. 

 Before admitting any gas into it, the bell glass was 

 exhausted by the mercurial pump to a pressure of 

 less than a millimetre, then filled with dry gas, and 

 again exhausted and recharged. The following table 

 shows the results obtained with spherical surfaces i'5 

 inch in diameter and 3 inches radius. 



Gas. 



Air 



Hydrogen 



Oxygen 



Carbonic anhydride 



Length of 

 Spark. 



Ratio of Length of Ratio of '_' J21ectric 

 of Spark. i Viscosity.*' 



Inch. 

 0*082 

 0-150 

 0'082 



0*077 



Referred 

 to Air. 



1*000 

 1*829 

 1*000 

 0*939 



Referred 

 tT Hydro- 

 gen. 



0-S47 



I'OOO 



0-547 



Referred 

 to Air. 



1*0060 



0-5467 

 I "0000 



0*513 1 1*0650 



Referred 

 to Hydro- 

 gen. 



■1*828 

 1*000 

 1*828 



1*949 



The appearance of the arc was different in different 

 gases, as will be seen in Fig. 15, where i represents the 

 arc in air ; this when examined with the microscope pre- 



Fig. 15. — The Arc in Different Gases 



sented an evidently stratified appearance, especially in 

 the barrel-shaped surrounding of the central bright 

 spindle ; the laminae were extremely close and seen with 

 very great difficulty, even with the revolving mirror of 

 the microscope ; with a moderate magnifier, a hand lens, 

 the barrel-shaped surrounding appeared as if shaded with 

 lines across it. 



The arc in hydrogen, with the point positive, is shown 

 in 2, Fig, 15 ; the central spindle was surrounded by a 

 beautiful blue halo like a glass shade illumined by fluo- 

 rescent light, and very brilliant on the disc. With four 

 megohms resistance in the circuit, the streaming dis- 

 charge from the point positive, at 0*502 inch distance, 

 was carmine in colour. This distinctive colour is a proof 

 that the gas in which the discharge takes place is the 

 carrier of electrification in the streamer discharge. The 

 appearance of the arc when the point is negative is shown 

 in 4, Fig. 1 5 ; it moved about very rapidly and formed a star- 

 like appearance on the positive disc ; when the point Avas 

 negative, before the jump of the spark, a very pale glass- 

 shade-like halo, of a saddened olive tint extended from 

 the point almost to the outer periphery of the disc. 



The arc in nitrogen was reddish violet, and in oxygen 

 it presented an appearance similar to that in air. The 

 arc in carbonic anhydride is shown in 3, Fig. 15. 



