178 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1925 



5(X),000; insulator flash-overs by lightning have occasionally indi- 

 cated voltages as high as 1,500,000 or more. 



A model was made to scale representing cloud and transmission 

 line for a certain observed condition. By means of the lightning 

 generator it was found that when a flash occurred from this model 

 cloud 1 per cent of its voltage was induced on the model line. But 

 it is known by observation that the voltage induced on an actual 

 line under similar conditions is sometimes of the order of 1,000,000. 

 If this is 1 per cent of the voltage of an actual lightning flash, the 

 voltage of the flash must be 100,000,000 volts (figl 13). ^This gives 



Actual 

 Line 



jTL ^Induced Voltage 

 ,-^^r T Measured 

 ' Distance Measured 



On an Actual Line the Voltage Induced by a Lightning Stroke to 

 Earth may he measured. It is also possible to deternnine the Distances 

 Indicated. Because of its Magnitude the Voltage 

 of the Lightning Stroke, However, cannot 

 Be Measured by Putting a Voltage Measuring 

 Device Directly Between Cloud and Earth. 



_^_ . -'Hf^g J For Determining the Voltage of the Actual Light- 



■^ ^-iydtage ^ ningitroke.the Following Method is Used. 

 ^ Distance'^ Measured 

 Known A Model Line and Cloud to Seals is Constructed 



in the laboratory. By Charging the Model 

 Cloud with the Lightning Generator, The Cloud 

 Voltage is Known. The Voltage Induced on the Line May be 

 Measured and Expressed as a flotio or in Per cent of 

 the Cloud Voltage. Knowing this Ratio and the Voltage 

 Induced on the Actual Line, the Voltage of the Actual Light- 

 ning. Stroke Can he Readily Calculated. For Heavy Strokes 

 Under Certain Conditions the Voltage of the Actual Light- 

 ning Stroke May be of the Order of 100,000,000 Volts. 



Fig. 13. — Method of measuring the voltage of an actual 

 lli,'htiilng stroke 



a voltage of 100,000 per foot of spark (330 kv./m.), which, con- 

 sidering the possible error, indicates that the needle gap-spark 

 curve may hold even at these extreme voltages. In fact, a value 

 somewhat lower than the 150,000 volts per foot (495 kv./m.) of the 

 laboratory would be expected, since the flash starts at the low-air 

 density of the high altitudes. Time required to discharge would also 

 indicate a reduced gradient. It has been tlie opinion of many 

 observers that the voltage of lightning is relatively low or that the 

 flash, once started, would extend great distances at low voltages by 

 some progressive action. Progressive action is readily understood 



