586 Lord Rayleigh on the Induction- Coil. 



In modern coils the greater part would appear to arise from 

 the positive and negative potentials at the ends of the coil as 

 opposed to the zero potential of the 'primary wire. The 

 capacity between the primary and secondary vires, considered 

 as poles of a condenser, can be calculated and in many cases 

 determined experimentally. The axial dimension of the 

 secondary of the coil above referred to is about 18 cm., and the 

 external diameter of the primary wire is about 5 cm., making 

 the area of each of the opposed surfaces 270 sq. cm. The 

 interval between the primary and secondary wires is '25 cm. ; 

 so that, taking the specific inductive capacity of the inter- 

 vening layer at 3, we get for the capacity in electrostatic 

 measure of the condenser so constituted 



■i X 3 x ~ = 258 cm.* 

 4-7T "25 



Only a fraction of this, however, is operative in the present 

 case. On the supposition of a coil constructed in numerous 

 sections, the potential in the middle will be zero, the same as 

 that of the primary wire, and will increase numerically 

 towards either end. The factor of reduction on this account 



will be \_* x 2 dci; or ^ so that we may take as the value of 



q in (6) about 23 cm. — probably rather an underestimate. 

 Calculating from these data, we get in (6) 



V = 2600. 



This is in electrostatic measure. The corresponding volts are 

 7*9 x 10 5 . If we reckon 33,000 volts to the cm., the spark- 

 length will stand at 24 Cm. The coil in question is supposed 

 to be capable of an 8 or 10 cm. spark, and doubtless was 

 capable when new. It is remarkable that the limit, fixed by 

 the iron and secondary capacity alone, should exceed so 

 moderately the actual capability of the coil. 



The limiting formula (6), in which neither the value of the 

 primary current nor the number of secondary windings 

 appears, is arrived at by supposing the iron to be magnetically 

 saturated. It illustrates, no doubt with much exaggeration, 

 the disadvantage of too great a length. If a be given, while 

 c varies, v and q are both proportional to c, so that V qc ^N. 

 And s/N oc c~ l nearly. In somewhat the same way the increase 



* Another coil by Apps, in which the insulation was sufficiently good 

 to allowthe application of electrostatic methods, was tested experimentally. 

 The capacity between primary and secondary wires was thus found to be 

 120 cm., less than the half of that calculated for the first coil. But in 

 this case an ebonite tube separated the two wires. 



