688 Prof. Fleming and Mr. Bichardson on the Effect of an 



spark-resistance for very short sparks. The good effect of 

 the air-blast is seen best when applied to short sparks, which 

 tends to show that it is of assistance in destroying the arcing 

 then occurring. This arc, however, is unable to persist with 

 longer gaps ; that is to say, it blows itself out, and accord- 

 ingly for spark-lengths of 3 mm. or upwards and when 

 using an ordinary 10 in. induction-coil with mercury break as 

 interruptor, the blast is of no special advantage. 



Another set of experiments was conducted, the object of 

 which was to ascertain the effect of the air-blast upon multiple 

 and upon very short spark-gaps. At one time some radio- 

 telegraphists were of opinion that an advantage was gained 

 by dividing up a spark-gap into smaller spark-gaps in series. 

 To test this opinion more carefully, a series of six spark-balls 

 were made, each consisting of a pair of brass balls 2 cm. in 

 diameter adjustable as to distance by a screw of 0'5 mm. 

 pitch, having a divided head. These spark-balls were arranged 

 so that they could be put in series with one another, the 

 series forming the spark-gap in an oscillatory circuit having 

 an inductance of 125,000 cms. and a capacity of 0*0045 mfd. 

 A number of glass jets were arranged so that an air-blast 

 could be directed against each pair of spark-balls across the 

 gap. In the oscillatory circuit a hot wire ammeter was 

 placed, so arranged that a reading could be taken of the 

 ammeter first with a single spark-gap of a definite length, 

 and then with a series of 2, 3, 4, 5, and 6 gaps, each adjusted 

 to be equal in length, the sum of them all being equal to that 

 of the single gap employed. Thus, for instance, the effect 

 of a single gap 0'5 mm. could be compared with that of five 

 gaps in series each 0*1 mm., both with the air-blast and 

 without the air-blast against the gaps. As in the case of 

 very short gaps, the phenomenon of multiple sparks exists ; 

 that is to say, each interruption of the induction-coil in the 

 primary circuit gives rise not merely to one discharge in the 

 oscillatory circuit, but to a series of discharges, because a 

 discharge takes place between the balls corresponding to the 

 length of the spark-gap employed, and whilst the electro- 

 motive force in the secondary circuit of the coil increases 

 and endures, during this time many sparks may take place. 



Whether this occurs or not can be determined by means 

 of a revolving mirror. If the image of an oscillatory dis- 

 charge produced by an induction-coil is examined in a 

 revolving mirror, then if there is only one spark correspond- 

 ing to each interruption of the primary coil, a series of 

 widely separated sharp images of the spark will be seen, but 



