May, 1905.] 



KNOWLEDGE & SCIENTIFIC NEWS. 



103 



book is in turn over each of the lower ones, in the alter- 

 nate positions shown in fig. 2. Whenever it reaches 

 either of these extreme positions, touch the book under- 

 neath it with the finger, taking care not to touch the 

 other book, nor the upper one. When the upper book is 

 in the intermediate position shown in fig. i it must be 

 touched, taking care to avoid the earthing of either of the 

 others. Repeat this cycle of movements twelve or twenty 

 times, and then, on lifting the glass plate with its book 

 away from the influence of the other two, a little spark may 

 be drawn from the penny on presenting a finger to it. After 

 drawing the spark replace the plate in the intermediate 

 position (fig. i),and on earthing the upper book again by 

 touching it as before, its charge will be restored. Con- 

 tinuing the movements, it will be observed that the charge 

 is a growing one, increasing at each cycle, being only 

 limited by leakage from the book corners and by the capa- 

 city of the book surfaces. The charge on the upper book 

 may be given up to a Leyden jar each time without loss 

 if the communication from it to the jar is always made 

 previous to its intermediate position, in which, when it is 

 earthed, its loss is made good by induction from the books 

 below. Using large books on the tumblers and, for the 

 upper one, a book of equally extensive surface, though thin 

 for the sake of lightness, quite a strong charge may soon 

 be accumulated in the Leyden jar. If the tumblers and 

 glass sheets are coated with shellac varnish they are less 

 liable to be affected by moisture, and the preliminary 

 warming is unnecessary unless the atmosphere is very 

 damp. The effect is enhanced by using rounded pieces 

 of board instead of the books, and the best effects of all 

 are produced if the conductors are of metal. Three 

 shallow cake tins, about eight inches in diameter, will 

 give very fine sparks, and the snap of the induced charge 

 quickly becomes audible at each successive earthing, the 

 spark of inflowing electricity becoming larger each time 

 until the limit of capacity is reached. The glass will 

 indeed soon become so highly charged that in moving it 

 the cake tin will adhere to it by attraction, while if the 

 lower tins are close together a spark will also frequently 

 ffy between them as the upper tin passes from one side 

 to the other. When this happens, however, the process 

 of accumulation is to some extent checked, and the proper 

 distance apart is the shortest distance at which such 

 cross-sparking does not occur. 



The experiment is really a modification of one which has 

 already been described and explained in " Knowledge " 

 (November, 1904). Each of the lower books receives an 

 infinitesimal charge by induction from the upper one, 

 which, without losing any of its own, is in its central 

 position enabled, when temporarily earthed by touching, 

 to take up an induced charge from the joint influence of 

 both the under ones. Its original charge is thus multi- 

 plied at each cycle. Where that original charge comes 

 from is, of course, a mystery. It is infinitesimal, but it is 

 there somehow, and may be either positive or negative — 

 sometimes one, sometimes the other. It appears that all 

 insulated bodies are at a slightly different potential from 

 earthed bodies, and though it may be difficult to say 

 exactly why they should be, the fact that they are is 

 sufficient to account for the apparent miracle of self- 

 excitement which characterises nearly all induction 

 machines. 



Perhaps one of the most interesting suggestions arising 

 from the experiment suggested above is in connection 

 with the phenomena of atmospheric electricity. The 

 effect of the movements of the books is so obviously sug- 

 gestive of the influence which cloud masses must exert 

 upon each other when one passes over two, with an 

 intervening space to separate them, that there is no resist- 



ing the conclusion that in the phenomena of the thunder- 

 storm we frequently witness on a large scale an almost 

 precisely similar experiment of Nature to that which we 

 have been performing on the dining-room table with the 

 three books to represent the cloud masses. Especially 

 suggestive is the flashing of the sparks from one of the 

 lower plates to the other, which occurs, as already men- 

 tioned, when they are placed very close together every 

 time the upper plate passes across them. Here undoubt- 

 edly we have the very counterpart of the phenomenon 

 often observed in a thunderstorm, when, drawn by the 

 influence of some upper layer of moving cloud, the light- 

 ning flash darts across from one charged cloud mass to 

 another in a lower stratum. 



Star MoLps. 



With this number we present the first of a series of Star 

 maps, which we hope will be found useful to our readers. 

 These charts of the heavens embody some new ideas of 

 design which, while causing the stars to be clearly 

 depicted as they appear in Nature, yet enable the 

 student at once to identify the individual stars and con- 

 stellations. 



One of the most difficult points to decide upon has 

 been the method of projection. It is, of course, im- 

 possible to represent all the objects on a spherical sur- 

 face, such as that which the heavens appear to be to our 

 eyes, in their exact relati\e positions on a flat piece of 

 paper. If, however, that piece of paper be cut up into a 

 number of small independent planes, each can more 

 accurately represent one portion of the sphere. So if 

 the surface of a globe were cut up into a hundred equal 

 parts, each of them would be practically flat, or if abso- 

 solutely flattened the position of the stars marked thereon 

 would not be greatly distorted. But such a series of very 

 small maps would be of comparatively little practical use. 

 The principal constellations would be split up into many 

 parts, and their general appearance lost. It is desirable 

 to form the maps in sheets as large as possible, both for 

 convenience of reference and for noting the relative 

 positions of stars and groups of stars. If we find one 

 object on a map we may wish to see in which direction 

 the various neighbouring stars lie, but this would be 

 almost impossible on very small maps. So for practical 

 work we require the maps to be as comprehensive as 

 possible. But any division of the globe into parts has, 

 to some extent, the objection that constellations and other 

 groupings are often divided. This may be overcome by 

 so arranging the maps that they overlap somewhat, and 

 the stars appearing near the borders of one may be 

 repeated on an adjacent map. 



Taking all these points into consideration, we believe 

 that no method can be better than that adopted by 

 Proctor in his " Star Atlas " (published in 18701, and it 

 seems appropriate that we should adopt the system 

 first introduced by the founder of "Knowledge." 

 We can but repeat the words of his Introduction. 

 "It is clear that, cceteris paribus, that plan is best 

 which represents the celestial sphere in the smallest 

 number of maps. Further the maps should be con- 

 venient in size but yet on a sufficiently large scale ; and 



