608 



SCIENCE 



[N. S. Vol. XXXII. No. 826 



vapor tension tables for water between 99° 

 and 101° C. to the attention of physicists. 



In conclusion, it should be emphasized that 

 we propose to use the boiling point method 

 only for getting differential results in gravity 

 and not for absolute results. All necessary 

 refinements are now to be introduced in the 

 future work. 



L. A. Bauer 



The Caenegie Institution 

 OF Washington 



SPECIAL ARTICLES 

 THE NATURE OF ELECTRIC DISCHARGE 



At a meeting of the Academy of Science of 

 St. Louis, on October 17, the vtriter presented 

 photographic plates which strongly confirm 

 conclusions reached in former papers.* Pin- 

 head terminals rest with their rounded heads 

 upon the film of a photographic plate. Their 

 distance from each other is about 7 cm. One 

 terminal is grounded in the yard outside of 

 the building. The other leads to a variable 

 spark-gap at the negative terminal of an 8- 

 plate influence machine, the positive terminal 

 being grounded on a water-pipe. With very 

 short spark-gaps, the passing of a single spark 

 produces discharge images immediately around 

 the pin-heads. Increasing the spark-length 

 enlarges the images, which are in the nature 

 of brush discharges. The negative glow 

 around the pin-head which communicates with 

 the negative terminal of the machine increases 

 very little in diameter, and the discharge lines 

 in it are radial. The discharge lines around 

 the grounded pin-head for short sparks follow 

 approximately the lines of force. With longer 

 sparks they are somewhat distorted, as if 

 beaten back by a blast from the opposite or 

 negative terminal. As has been suggested in 

 the papers referred to these discharge lines in 

 the " positive column " are drainage lines, 

 along which Franklin's fluid is being con- 

 ducted into the positive or grounded terminal. 

 The portions of the air molecules which con- 

 stitute the stepping stones for the negative 

 corpuscles are urged in a direction opposite 



' Trans. Acad, of Sc. of St. Louis, XIX., Nos. 

 I and 4. 



to that in which the negative discharge is 

 flowing, thus promoting the lengthening of the 

 drainage lines. Many hundreds of plates have 

 been exposed in an attempt to adjust the 

 spark-gap so that these drainage lines would 

 end just outside of the negative glow without 

 reaching it. In this way the length of these 

 lines may be gradually increased until they 

 approach the dark space around the negative 

 glow. This dark space is a region where con- 

 vection of atoms which have been super- 

 charged within the negative glow are urged 

 by convection away from the negative ter- 

 minal. If the drainage lines reach this con- 

 vection region, they cross it and reach the 

 negative glow. It has thus far been found 

 impossible to have them end within this Fara- 

 day dark space. If the spark-gap at the 

 machine is so adjusted that only one or two 

 di-ainage lines reach the negative glow, these 

 lines will unite end on with the radial dis- 

 charge lines of the negative glow. At the 

 same time there is a distortion in the lines at 

 and near their iinion, which reveals the com- 

 motion produced by the opposing " electric 

 winds." 



If now the spark-gap at the machine be 

 slightly increased, other drainage lines reach 

 the negative glow. They cross its radial dis- 

 charge lines, and even extend beyond the nega- 

 tive terminal. In a few cases the entire area 

 of the negative glow is traversed by these 

 drainage lines. It is evident that we have 

 here the same conditions that Goldstein found 

 in the vacuum tube. These drainage lines 

 are the canal rays of the vacuum tube. 



This explanation of the nature of electric 

 discharge enables us to understand why the 

 positive column in a vacuum tube follows 

 the tube in all of its windings and bends. It 

 is not a convection column, but a drainage 

 column. It is a conduction column. The 

 conditions are different from those in a copper 

 wire, in that the parts of the atoms which 

 constitute the conductor are in gaseous form, 

 and are capable of yielding to the force which 

 urges them in a direction opposite to that in 

 which the negative corpuscles are being urged. 

 Francis E. Nipher 



