ON PH0T05EAPHT OF METEOROLOGICAL PHENOMENA. 137 



the contrary seems often to be the case. Again, if the former hypothesis 

 were true, the position occupied by the reflected light could be ascer- 

 tained by considering the direction of the incident light. Fact here 

 disagrees with theory. 



The evidence at present obtainable therefore points to the conclusion 

 that a bright lightning flash may often take the form of a long sinuous 

 ribbon, whose sectional thickness is very different in two directions 

 normal to each other. Some of the appearances noticed also indicate that 

 the greater thickness throughout all the parts of a given flash lies in one 

 and the same direction, and the variations in its apparent direction are- 

 mei'ely an effect of perspective. 



This structure must be carefully distinguished from another, in which 

 several distinct flashes follow precisely similar paths side by side. Some- 

 times the bright flashes (which may or may not show the ribbon shape 

 proper) are connected by a less brilliant luminosity, which converts the 

 whole phenomenon into a very broad ribbon. Photographs of this class 

 are exemplified by Nos. 4B, 5 B. The flash represented in Dr. Hoffert's 

 photograph is evidently one of the same order, and the curious smudges 

 which cross the plate must doubtless be due to the above-mentioned 

 fainter light. Clearly we have here to deal with intermittent dischai'ges,. 

 a number of discharges following each other along the same or closely 

 contiguous paths. In some cases photographs of this kind show redupli- 

 cated images of buildings corresponding fairly well with the images of 

 the component parts of the dischai-ge. In such a case there seems little 

 room for doubt that the flashes followed the same path or paths only a 

 very short distance apart. 



The secretary to your committee, however, secured the photograph 

 No. 4 B on June 25. In this case the camera was certainly not moved. 

 The flash, like many others, appeared multiple to the naked eye, but as 

 the motion of the eyeball might have produced tliat eff'ect, although the 

 flashes formed the same path, little weight can be laid on that argu- 

 ment. Indeed, the fact that the camera was standing still and quite un- 

 touched is sufficient to prove that flashes of such a nature do occur. It 

 is really a rapid and almost simultaneous volley of flashes connected 

 partly hy a less vivid discharge which obliquely links the brighter lines. 

 There is also evident a sort of half-twist of one part of the flash around 

 another part. 



In order to elucidate the unexpected facts brought to light in the 

 numerous photographs belonging to the Royal Meteorological Society a 

 number of experiments have been made by your secretary upon electric 

 sparks obtained from an induction machine. As these tend to throw 

 some light upon the questions in hand, a brief account of them may not 

 be out of place. 



First remove the small Leyden jars from a Voss or Wimshurst 

 machine. The discharge is then pink in colour, of slight brilliancy, and 

 strongly resembles the brush discharge. If the knobs are brought 

 near each other the dischai'ge passes along several lines, which arrange 

 themselves side by side in a plane at right angles to the direction of 

 discharge. 



If now the condensers are introduced in the ordinary position, the 

 spark at once becomes more brilliant, and the pink tinge disappears. 

 This spark obtained from the ordinary size of condenser appears to b& 

 precisely the same as the commoner varieties of lightning. If larger 



