Intelligence and Miscellaneous Articles. 



517 



"We have greatly varied the conditions of the experiment, and 

 repeated the observations a great number of times. 



We collate in this Table the results we have obtained with five 

 different circles*. 



Diameter 



lm. 



stout 



0-75 



0-50 



0-35 



0-35 



fine 



wire, 



2 mm. d. 



0-25 



0-25 



0-20 



0-20 



010 



of the 



stout 



stout 



stout 



stout 



fine 



stout 



fine 



stout 



circle D. 



1cm. (7. 



wire. 



wire. 



wire. 



wire. 



wire. 



wire. 



wire. 



wire. 



1st Loop 



211 



1-60 



Ill 



0-76 



0-75 



0-46 



0-54 



0-39 



0-42 



0-21 



1st Node 



4-14 



3-01 





1-49 



1-51 



0-94 



147 



0-80 



0-93 



0-41 



2nd Loop . . . 



>> 



,, 





230 



237 



1-63 



1-89 



1-24 



1-55 



0-59 



2nd Node ... 







} , 



3-04 



340 



215 



2-40 



1-69 



2-05 



079 



3rd Loop . . . 





>) 



,, 



, 



,, 



2-71 



2-94 





2-46 



0-96 



3rd Node ... 







J5 



„ 1 



} , 



314? 



„ 









£\Air 



2-03 



1-41 



Ml 



0-76 



0-80 



„ 



0-60 



0-43 



Ool 



019 



JXWire ... 



1-92 



1-48 



0-98 



0-73 ; 



,, 



fJ 



0-56 





0-45 



,, 



2 D 



2-00 



1-50 



1-00 



0-70 



0-70 



" 



0-50 



0-40 



0-40 



0-20 





We give in this Table the means of the measurements obtained 

 with each circle, and for the sake of brevity we have not separated 

 those made with primary exciters of different dimensions, as these 

 did not present systematic differences. In these latter experiments 

 in air, as in our previous researches along wires, we have in part 

 established that a circular resonator always gives tJie same wave- 

 length, even when the dimensions of the primary are varied within 

 certain limits f. Then, again, is observed what we have called 

 multiple resonance. 



In the case of larger wave-length circles of 1 m. and of 0*75, 

 we can scarcely determine with any precision more than a node 

 and a loop, besides the node on the mirror itself. With smaller 

 circles for jwhich the dimensions of the mirrors are somewhat more 

 suited we can easily determine three loops and three nodes, in- 

 cluding that of the mirror. The equidistance of the loops and 

 nodes is, as will be seen, pretty satisfactory ±. 



The most important result of our research follows from the com- 

 parison of the figures of the three last lines of the Table, which 

 show that the wave-length obtained for each circle in the case of 



* The delicacy of the micrometric screw is a very important element 

 in this kind of research, especially with the small circles, which only give 

 very feeble sparks. In the latter case we used a screw giving the j^ of 

 a millim. 



t As the intensity of the spaik of the resonator is feebler in this expe- 

 riment than in that "of the wires, and as it diminishes, moreover, much 

 more rapidly as we move away from the primary, the observation is 

 altogether less precise than with wires. In order to work in good con- 

 ditions the primary should have dimensions suited to the diameter of the 

 resonator. The limits within which we can conveniently observe multiple 

 resonance are less extended than in the case of wires, but vary up to 

 double or even more. 



\ M. Hertz places the first node at a certain distance behind the 

 mirror ; this perturbation at the end does not seem to follow from our 

 experiments. As will be seen from the Table, the first quarter of a wave- 

 length presents no systematic difference from the others. 



