142 REPORT— 1893. 



tiou of a provisional cloud atlas, or one section of it, under the following 

 conditions. Divide all clouds into the three great groups, Cumulus, 

 Stratus, and Cirrus. Publish volumes dealing with each of these great 

 groups, not naming the subordinate varieties, but assigning merely num- 

 bers. Thus, supposing there are ten varieties of Cirrus — call them 

 No. 1, No. 2, &c. ; then, if there are ten varieties of cirro-cumulus, let 

 these be numbered from eleven upwards. 



It seems that if such an atlas were distributed to a number of ob- 

 servers who are in the habit of making eye-estimations of the quantity 

 of cloud, it would be quite easy for them to record also the numbers of 

 the respective types of cloud visible. Since these observations would be 

 made by meteorologists, and at the same time as records of temperature, 

 pressure, &c., the results could not fail to be of real importance. 



Again, a meteorologist armed with such an atlas would be able to 

 note changes of form from one type to another almost as well as the actual 

 photographer. 



Lightning Photographs. 



Not many new photographs of lightning have been received, but 

 they all agree with the others in the Committee's collection in showing 

 what has been called the narrow ribbon structure. There has not yet 

 been any opportunity of ascertaining whether this structure is shown in 

 negatives on paper, but it is visible in negatives taken on thin films. 

 This fact confirms the opinion already expressed by your Committee, 

 that it represents the true form of a lightning flash. Moreover, it cannot 

 be caused by reflection from the back of the plate, because if so it would 

 be most evident in the brighter parts of the flash, whereas it is most 

 evident in the fainter ; also, it would be more pronounced in the margins 

 of the plate than in the centre, and the apparent orientation of the 

 ribbon would vary according to the position on the plate. None of these 

 things are noticeable. The major thickness of the ribbon seems to set 

 itself in a particular direction, which is constant for all parts of a 

 branched or other flash, whatever may be the position of the image on 

 the plate. It is also not a whit more obvious in the margins of the 

 plate than in the centre. Lastly, it is almost invariably shown more or 

 less plainly ; why, then, should it be supposed to be due to some error of 

 observation ? 



It has lately been suggested that it is produced by marginal deforma- 

 tion of the image. Let us put aside for a moment the fact that the 

 phenomenon is not marginal at all. Now, if a lens be used which will 

 not cover the plate properly, so that the margins are out of focus, or if 

 the camera be purposely put out of focus, it is quite true that the 

 image of an electric spark may be expanded into a broad ribbon. But 

 this is characterised by both the margins of the ribbon being brighter than 

 the centre, while in the true narrow ribbon structure, as shown by light- 

 ning, the whole is equally bright, or one margin is bright and the other 

 the faintest part of the image. The explanation is clearly incorrect. 



It may be useful here to draw a definite distinction between a light- 

 ning ' flash ' and a lightning ' discharge.' Flashes last only a short time, 

 a mere fraction of a second, though probably a considerably larger frac- 

 tion than was at one time supposed. The eye is not conscious of any 

 variations of brilliancy during the flash, and a camera moving with con- 

 siderable velocity does not resolve it into a number of components. 



