6 Prof. E. Edlund's Researches on the Passage 



ments made, in order to economize space I shall cite only 

 those which are absolutely necessary to furnish a complete 

 demonstration*. Fig. 1. 



In fig, 1, ah represents the 

 disk of the electrophorus, a x and 

 b x the two receiving combs, of 

 which a x is in metallic connexion 

 with the point k, and b x with the 

 brass ball e. In the vicinity of 

 this latter another brass ball,/, 

 is placed, from which an elec- 

 trode goes to the point h. The 

 space in which the air is to be 

 rarefied is represented by c d, 

 and is connected with the points 

 h and k. From these points also 

 electrodes go to the galvano- 

 meter g, in front of which the 



wires are joined by a bridge n, which presents a suitable 

 resistance: the purpose of this bridge will be seen further on. 

 I is a copper wire connected to earth, and carries away the 

 static electricity which has possibly been left in the galvano- 

 meter after the discharges, and might produce an electroscopic 

 effect upon the magnetized needle. 



When the disk of the electrophorus is put into rotation, the 

 sparks spring between the balls e and/, and the magnetized 

 needle of the galvanometer g is deflected. If the rotation be 

 continued for a sufficient time, the deflection will become 

 constant. The arrival of that moment, however, was not 

 always waited for; but the amount of the deflection was de- 

 termined by observing the degrees of the scale to which the 

 needle returned in its oscillations ; then a mean of those de- 

 grees was taken. Several such means were taken for each 

 determination. If (as indicated in the figure by an arrow) 

 the positive current passes from the receiving comb a 1} this 

 current divides at k, after which a portion of it traverses the 

 space of rarefied air, while, on the other hand, a certain quan- 

 tity passes through the bridge n, and the remainder through 

 the helix of the galvanometer. But when the conduction is 

 interrupted between the point k and the rarefied air-space, the 

 entire current passes to q, and is there divided between the 



* The observations were carried out, under my direction, by MM. S. 

 Arrhenius and 0. Mebius, Candidates in Philosophy of the University of 

 Upsal, and Th. Homen, Candidate in Philosophy of the University of 

 Helsingfors (Finland). Usually each observation was repeated three 

 times, which permitted the observers to control each other. 



