) 



152 



SCIENCE 



[N. S. Vol. XXXIII. No. 839 



that it is in gaseous condition. The apparent 

 discharge into the air from the positive ter- 

 minal was shown by phenomena described in 

 my paper of February 18, 1910/ to be an in- 

 flow of negative electricity to that terminal. 

 The electric fluid is thus drained from a col- 

 umn of air which begins at the positive ter- 

 minal and extends towards the negative or 

 compression terminal. After this drainage 

 has been brought about, the air within this 

 column is in a very different condition from 

 that of air ionized by X-rays. The average 

 charge per molecule of air is then less than 

 the normal amount. By placing an insulated 

 sheet of copper between the terminals, the 

 drainage or conduction column is prevented 

 from reaching the mass of supercharged air 

 in front of the negative terminal. Loud dis- 

 ruptive discharges will pass when the plate is 

 removed, and they cease at once when the 

 plate is placed midway between the terminals. 

 On moving the copper plate nearly to the nega- 

 tive glow, the drainage column follows it, and 

 a torrent of sparks will pass. 



The so-called positive ions do not emerge 

 from the wire at the positive terminal, and 

 they do not enter the wire at the negative 

 terminal. When this air is nearly all removed 

 from the discharge gap, as is done in the 

 Crookes tube, the cathode discharge still con- 

 tinues. The drainage column has disappeared 

 at the anode. When the air is partially re- 

 moved, thus increasing the mean free path, 

 the drainage column may have a great length, 

 and it follows all of the windings and bends 

 of the tube. It would not do this if it were 

 an outward discharge like that from the nega- 

 tive terminal. 



This drainage column constitutes the canal 

 rays, in a tube where the gas has been partly 

 exhausted, so that the mean free path has 

 been increased. In a paper to be published by 

 the Academy of Science of St. Louis, phe- 

 nomena of canal rays in air of ordinary pres- 

 sure will be presented, in connection with 

 phenomena involved in vibrations imposed 

 upon the air column. 



^ Trans. Acad, of So. of St. Louis, Vol. XIX., 

 No. 1, Plates II. to VIII. 



If any trace of a positive fluid capable of 

 moving through a wire as an electric current 

 must move could be found, and if the two- 

 fluid theory could furnish an adequate and ra- 

 tional explanation of these and other phe- 

 nomena, there would be no objection to its use 

 in an exposition of the subject, as has been the 

 custom heretofore. 



In 1895 the writer showed that the velocity 

 of flow of the electric fluid in a pumping serv- 

 ice then discussed, must be very great. Imag- 

 ine two spheres having radii equal to that of 

 the earth. Suppose electricity to be pumped 

 from one into the other, until their difference 

 in potential is 50 volts. Connect them by 

 means of a 50-volt one-ampere lamp. It 

 would begin to glow with normal brightness. 

 In order to keep the voltage constant the radii 

 of the two spheres must be diminished with a 

 uniform velocity of more than half the veloc- 

 ity of light. This store of electricity would 

 maintain this lamp at normal candle power 

 for 0.035 second. The operation must be re- 

 peated 28 times a second. The time for one 

 stroke of the piston is in seconds, t:=Br, 

 where r is the radius of the spheres in cm. and 

 R is the resistance of the lamp in electrostatic 

 units. If the sphere from which the electric 

 fluid were pumped, by some adequate means, 

 had an infinite capacity, the other sphere must 

 then be charged to a potential of 50 volts, and 

 the above results would remain unchanged. 

 The surface of the charged sphere would then 

 be the piston of an electrical pump, and its 

 velocity would be as given above. All of this 

 store of electricity must pass through the 

 lamp in 0.035 second. 



We may learn much about the constitution 

 of matter by a study of what are called the 

 positive ions when an electric discharge is- 

 passed through a column of gas. But nothing 

 has yet been learned to indicate that these 

 ions play any part in a lamp circuit, except 

 that they then constitute the solid conductor. 



Wheatstone's work now shows us that when 

 half a mile of copper wire is placed centrally 

 in the spark gap, we have compression and 

 rarefaction waves in Franklin's fluid, which 

 meet at or near the middle of the wire. The- 



