Decembeb 17, 1909] 



SCIENCE 



865 



of the sparks in space from records so pro- 

 duced. If this were done as it were 

 stereoscopically or stereographically from 

 two or more separated observing or record- 

 ing places, the records would contain the 

 necessary data for the reconstruction of the 

 spark and its branches in space. 



From the above considerations an at- 

 tempt to determine the distance of a light- 

 ning stroke to earth by counting seconds 

 elapsing between the flash and the first 

 thunder and allowing five seconds to a 

 mile approximately is seen to be futile. 

 Should one of the cloud ramifications or 

 branches of the great tree-like discharge ex- 

 tend in the cloud overhead with relation to 

 the observer, and that part of the dis- 

 charge be nearer to him than any other he 

 will first hear a receding rumble above 

 him, followed it may be by a heavy ex- 

 plosion from the main or approximately 

 vertical spark between cloud and earth and 

 from the parts of which his distance is 

 nearly the same. This louder explosion 

 will then be followed generally by a pro- 

 longed rumble of diminLshing loudness 

 which is the sound coming from the rami- 

 fications which lead farther to the distant 

 parts of the cloud. Manifestly the count- 

 ing of time should be between the flash 

 and the heavy explosive sound due to the 

 vertical part of the flash. 



Bearing in mind that over the extent of 

 cloud the charged water particles may be 

 said to be waiting for a chance to dis- 

 charge to earth, it is not surprising that 

 any path -which has been opened or broken 

 down by di.sruption of the insulating layer 

 of air should serve for the discharge of an 

 extended body of cloud. The heated vapor 

 or gas in the path of the discharge is a 

 relatively good conductor of electricity 

 serving to connect the cloud mass to the 

 earth below. The significance of this is 

 imderstood when it is known that many 



lightning discharges are multiple. In- 

 stead of a single discharge they consist of 

 a number rapidly following one another 

 through the path or spark streak opened to 

 them by the first discharge. This first dis- 

 charge opens the way or overcomes the 

 insulating barrier to the discharge of por- 

 tions of the cloud mass, which, on account 

 of remoteness or lower potential, could not 

 themselves have caused the breakdown. 

 The.se repeated or multiple flashes are ex- 

 ceedingly dangerous, both to life and prop- 

 erty. The first discharge may reduce wood 

 to splinters and the subsequent ones set it 

 on fire. The time interval between the 

 successive discharges in such a multiple 

 flash is quite variable and may be long 

 enough to be easily perceptible by the eye. 

 The multiple character is easily disclosed 

 by the image in a revolving mirror. If a 

 strong wind be blowing at the time of 

 such a multiple flash, the hot gas conduct- 

 ing the discharges may be displaced later- 

 ally in the direction of the wind with the 

 result of spreading out the discharges into 

 a ribbon more or lef?s broad. Photographs 

 of these ribbon flashes show their true char- 

 acter plainly; each separate discharge ap- 

 pearing as a streak of light parallel to the 

 others and at varjnng distances apart. In 

 fact parallel discharges of exactly the same 

 contour are sometimes observed many feet 

 apart. Here the hot gas of the first dis- 

 charge has evidently been shifted by the 

 wind over a considerable space before the 

 second and subsequent discharges took 

 place. Heavy rain seems to weaken the 

 air and help to precipitate a discharge. 

 Prom the fact that strokes of lightning are 

 often followed by increased fall of rain 

 within a few seconds it is a prevalent idea 

 that the increased downpour is caused by 

 the discharge. In reality the reverse is the 

 case, for just when a gush of rain has 

 reached from the cloud down to within a 



