60 MM. Strouhal and Barus on a Simple 



It is moreover immaterial whether we employ in the results 

 Fresnel's representation of the surface of the waves or that 

 deduced from Cauchy's theorem : the latter only is theoreti- 

 cally accurate ; but I have before demonstrated that, on de- 

 termining by experiment the constants which enter into it, it 

 coincides with Fresnel's, and that with an exactness exceeding 

 that of the observations. 



It is remarkable that the transformation indicated above for 

 Fourier's formula has no existence for a function of two vari- 

 ables. This fact is connected with another, viz. that there 

 cannot exist a limited cylindrical wave, at least interiorly. 



Supposing the initial disturbance circumscribed within an 

 indefinite vertical cylinder, admitting moreover that the ex- 

 cursion and the velocity are at that instant the same for all 

 points situated on one and the same vertical, this will certainly 

 be the case during the whole motion ; and this will have the 

 form of a vertical cylindrical wave propagating itself horizon- 

 tally in all directions ; but the whole of the points interior to 

 the cylinder will remain indefinitely in motion ; so that it 

 cannot be resolved into limited plane waves. 



IX. On a Simple Method of Galvanic Calibration of a Wire. 

 By V. Strouhal and C. Barus *. 



THE methods which are customarily used for the calibra- 

 tion of a wire (as, for example, in the Kirchhoff-Wheat- 

 stone bridge-combination) undoubtedly labour under the dis- 

 advantage that for carrying them out resistances are required 

 beforehand which have been equalized in another Way under 

 certain simple conditions. In this way the accuracy of the 

 desired result is made dependent on that of a previous equali- 

 zation, whereby the errors to be expected in the calibre of the 

 wire are placed in doubt by unavoidable errors in the equali- 

 zation of the resistances, and so much the more the smaller 

 they happen to be. Hence the importance of the problem for 

 precise determinations by the otherwise so convenient bridge 

 method constantly justifies the endeavour to get rid of this in- 

 convenience as completely as possible, and, in addition, by 

 the simplicity of the means propounded, to make the execu- 

 tion to the utmost convenient and facile. 



The method described in the following, which we have re- 

 peatedly employed with advantage, is perfectly analogous to 

 that which is usually applied, in the well-known manner, to 



* Translated from "Wiedemann's Annalen, 1880, No. 6, vol. X. pp. 326- 

 330. 



