meridian . . . with the needle between the wim 

 I. .m h single i ircuil . gave .1 deflei don of 45 .... 

 When tw <» were connected the deflection was <>" 3 , and 

 when Snail) all three were pui in magnetic operation, 

 the deflection grew to only 70 . It appears clearly 

 in. in 1I1U that the angle ol deflection is not in a simple 

 ratio with the magnetii force ai ting on the needle .... 



Neither Poggendorf nor Schweigger seems to have 

 ruled "in. .hi logical grounds alone, the possibility of 

 deflections greater than '•n J . with tin- loop-plane in 

 the magnetic meridian, though Poggendorf does add 

 a vague note that if the needle deflected too far it 

 would encounter forces ol the opposing sign. 



Poggendorf experimented with the size of the cir- 

 cuit wires, finding that larger wires led to gn 

 deflections. He noted thai the size of the cell plates 

 and the nature of the cell's moist conductors would 

 certainly have a great effect, but that to investigate 

 these in detail would take undue time, and he there- 

 fore proposed to keep this part of the apparatus 

 constant, using one pair of zinc and copper plates 3.6 

 inches in diameter, separated by cloth soaked in 

 ammonium-chloride solution. 



Poggendorfs principal quantitative stud) of his 

 magnetic condenser used 13 identical coils, each 

 with 100 turns. In order that the turns should all be 

 at approximately the same distance from the needle. 

 the coils wen- wound of the finesl brass wire thai eon Id 

 be silk-insulated, the wire diameter being 0.02 lines. 

 On adding coils one at a time across the cell (i.e., 

 connecting them in parallel), the deflections were as 

 follows 



Turns too qoo 300 400 500 600 700 



Deflection in 

 degrees 4-, 50 55 59-60 62 63 64 



[urns 800 qoo 1000 1100 1200 1300 



Deflection in 

 degrees 65 65] 66 66 66 66 



Adding some coils with fewer turns, and connecting 

 various combinations "as a continuum" (i.e.. in series), 

 the deflections using the same cell were: 



Turns 1 5 10 25 50 75 100 200 



Deflection in 

 degrees 10 22 27 30 15 lo lo 40 40 



Turns 300 400 500 600 700 800 900 1000 



Defection in 

 degrees 40 40 41 40 40 40 40 w 



Making a few coils from wire with \-line diameter, 

 the deflections, again using the same cell wei 



Since the needle used in these experiments was 

 almost as long as the inside clearance ol the coils, 

 no simple tangent law can be applied, and it is nol 

 possible to disco vet an equivalent circuit in modern 

 terms. However, the constancy of the deflei 

 for large numbers of turns in eat h case indicates thai 



the cell volta 1 n isl 1 istant, 



and a rough estimate suggests thai the cell resistance 



was comparable to the resist 1 one ol the 100- 



turn coils of fine wire. Such a value means thai c< 11 

 resistance limited the maximum deflections for the 

 parallel-connected multipliers, while coil resistance 

 fixed the limit in the series case. 



For all of these reasons, it was impossible thai any 

 useful functional law could be obtained from the data. 



I'oggendorl concluded onl\ that "the amplifying 

 powei of the condenser does not increase without 

 limit, but has a maximum value dependent on the 

 conditions of plate area and wire size." He added 

 two other significant comments derived from various 

 observations, thai the basic Oersted phenomenon is 

 independent of the earth's magnetism, and that the 

 phenomenon is localized, i.e., is nol affet ted b) distant 

 parts of the circuit. 



Onl) a small fraction oi Poggendorfs paper is 



devoted to elucidating the properties of the condenser. 



A similar amount is concerned with refuting \ 

 proposals, such as those of Berzelius and lam. in. 

 about distributions of magnetic polarit) in a conduct- 

 ing wire to account foi ' Oersted's results. More than 

 half of the paper describes results obtained by using 

 the condenser to compare conductivities and cell 

 polarities under conditions where no effeel had 

 previousl) been detectable. Notable is the observa- 

 tion of needle delleclions in i ireuils whose ec 11 meet ing 



wires are interrupted b) pieces of graphite, man 



dioxide, various sulphur compounds. ,1, . mi 

 which had pieviouslv been considered as insulators 

 in g ilv mil circuits. Poggendorf gives these the name 

 ol ■semi-conductor" (halb-Lt 



Cumming's first mention of the multiplier phenom- 

 enon, in his paper of April 2, 1921, ■'•' is quite casual, 

 and describes only a one-turn construction. 1 le speaks 

 lirsi of single-turn ring of thick, brass wire, and after 



PAPER 38: EARLIEST I I I ( 1 ki iM \( ,\l I 1( INSTRUMENTS 



133 



