188 Prof. W. E. Ayrton and Mr. T. Mather on the 



dipping into the mercury-cups, and, when all are in series, 

 the resistance of the 72 metres of strip at 15° 0. is 2*932 

 ohms. The resistance of the arrangement does not alter by 

 more than one tenth per cent, when a current of 15 amperes 

 is passing through each strip. 



The wooden frame which carries this series of platinoid 

 strips hangs on paraffined ebonite pegs attached to one of the 

 laboratory walls, so that it is well insulated. 



The inductance of the set of strips, even when all are in 

 series, is so small that we have not been able to measure it 

 even with the secohmmeter. Some of the tests have given 

 indications of a small negative result, which, if true, would 

 mean that the capacity-effect slightly overbalanced the induct- 

 ance ; but a calculation, which we have made, appears to 

 show that with the dimensions in question such a result is 

 impossible. We can therefore only conclude that this resist- 

 ance-frame, which was constructed by two of our former 

 students, Messrs. C. Gr. Lamb and E. W. Smith, fulfils the 

 object for which it was intended so well that the inductance 

 cannot be detected with certainty by any test that we have 

 hitherto tried. We therefore have decided to duplicate the 

 arrangement, the wooden framework (only one half of which 

 is seen in fig. 1) having been constructed large enough to 

 hold a second set of strips. 



Another method of constructing non-inductive resistances 

 for large currents, which has been in use at the Central 

 Institution for the past eighteen months, is illustrated in 

 fig. 2, and consists in winding two bare platinoid wires of 

 equal length and thickness into two spirals, one right-handed, 

 the other left-handed, the diameters of the two spirals differ- 

 ing slightly from each other so that one spiral can be placed 

 inside the other. They are then connected up in parallel, 

 so that when a current is sent through them it circulates 

 clockwise round one spiral and counterclockwise round the 

 other, the magnetic effects of the two thus tending to neu- 

 tralize one another and to produce a combination with small 

 inductance. 



The inductance of coils constructed in this way is almost as 

 small as if the wires were doubly wound like an ordinary 

 resistance-coil, but they possess the great advantage that 

 parts differing much in potential are not close to one another. 

 The wires therefore require no insulating covering, for no 

 harm will occur if one spiral accidentally touches the other, 

 provided that reasonable care has been taken to space the 

 convolutions fairly uniformly. Further, as the cooling sur- 

 face, for a given total cross section of the conductor, is much 

 greater for two concentric spirals than for a single spiral of 



