746 POPULAR SCIENCE MONTHLY. 



running to the ground. There was some light, but no incandes- 

 cence of the filament. It was more in the nature of a creeping 

 of the charge over the outer glass surface of the lamp. Stinging 

 sparks were felt whenever the kite wire was touched. The storm 

 gradually passed over, the lightning being vivid and frequent in 

 the west and north, and, as we learned next day, doing consider- 

 able damage. The nearest flash to the hill, however, as well as 

 we could determine by the interval between thunder and flash, 



Harokave Kite in Aik. Same kite as in preceding cut. 



was forty-five hundred feet away, so that the discharge which the 

 observer felt while loosening the wire must have been a sympa- 

 thetic one. We obtained a photograph of the prime discharge, 

 and very curiously this shows a remarkable change of direction. 



This year, in some interesting experiments made on the roof 

 of the Mills Building at San Francisco, it was noticed that the 

 roof, which has a covering of bitumen, was a good insulator. 

 Ordinarily one may touch the reel on which the kite wire is 

 wound without being shocked, but if a wire be connected with 

 the ventilating pipes running to the ground there are small 

 sparks. Introducing a condenser in the circuit, the intensity of 

 the spark is increased. It only remains to construct an appro- 

 priate coil of the kite wire and place within it another inde- 

 pendent coil. In the outer coil a quick circuit breaker may be 

 placed, and theoretically at least we shall transform down the 



