Mabch 1, 1900.] 



KNOWLEDGE. 



53 



though the fig^iircs, as a whole, ai'e lai'ge, as in other 

 eases where frouds arc produced. 



As the ablest electriciaus axe not yet agreed as to what 

 is the real nature of electricity. I shall not presume to 

 cuter on a discussion involving this question, but if any 

 readei^s iutei'csted in such speculations will refer to 

 Lord Armstrong's " Electric Movement in Air and 

 Water. " they will probably be struck with how readily 

 the theoiy broached by him lends itself to an explana- 

 tion of the experiments above described. The eonccn- 

 trative cluu-aeter of the neg<itivo current, as opposed 

 to the dispersive chai-acter of tho positive, which is 

 involved in his theory, fits in well both with the more 

 penetrative power of the former, and also with its being 

 less eiisily diverted by resistances when moving in the 

 plane of the paper, and thus forming smoother curves 

 when the homogeneous nature of the support for the 

 sensitive film admits of this. 



The method of receiving the spark discharge on a 

 photographic plate may also be employed to show the 

 oscillatory character of the discharge of a Leyden jar. 

 As is well known, this discharge is not one simple 

 operation, but if we suppose the inside to be charged 

 positively, then, when tho jar is discharged, this positive 

 chai-ge does not merely leave the inside, but surges over 

 to the outside and charges it positively, instead of 

 negatively, as it was before. This new positive charge 

 on the outside then surges back to the inside, and this 



1 \ 



Fig. 6. — Negative lying belon' Fig. 5. 



oscillatory process is repeated with inconceivable 

 rapidity until the oscillations gradually die away. 

 Fig. 7 is taken from a plate which faced the negative 

 terminal, and yet it wnll be seen that the positive mark- 

 ings from the oscillation caused by the Leyden jars 

 extend nearly as far as the negative ones. Even when 

 there is no Leyden jar in the circuit, there may be 

 oscillations iu the discharge of sufficient strength to 

 show themselves on the plates. On closely examining 

 Fig. 2, there will be seen small positive twigs projecting 

 from some of the main lines, while negative frouds may 

 be traced, though less easily, on Fig. 1. I have found, 

 in general, that the superior efficiency of the negative 

 discharge assei-ts itself here also, so that the effects of 

 its return as a positive charge are more easily recog- 

 nized in the negative plates than are those of the negativ2 

 return swing of the positive charge on tho positive 

 plates. When the Wehnelt interrupter was used in- 



stead of the ordinary contaio muker, then, eveutlmugii 

 there was no Leyden jar introduced, tho glass plates 

 showed the central rayed figures, in addition to the 

 normal positive branches and negative fronds. As I 



FlO. 7. — Glass negative with two Leydeu jars, showing oscillations. 



had cut out the condenser when using the Wehnelt 

 interrupter, these rayed figures were probably produced 

 by the spark at " make," and the normal ones by the 

 spark at " break." If the spark be allowed to pass 

 freely between the terminals of the coil, with no plate 

 or other obstacle interposed, the portion which produces 

 the rayed figures will be seen to be of a dull magenta 

 colour, while the other portion has the usual white 

 colour. Oscillations were well shown on some of the 

 images produced with the Wehnelt interrupter. It 

 will be seen that these experiments can easily be re- 

 produced, and any one who repeats them under varying 

 conditions will find that they exhibit a number of 

 details which it would be of more interest for readers to 

 work out for themselves than to have them minutely 

 described for them. 



POLARITY IN MAGIC SQUARES.-II. 



By E. D. Little 



IV. Natural Chain.— The numbers 12 3-156789 

 may of course be arranged in the form of a square in 

 such a way that every two successive numbers in tho 

 series shall occupy contiguous positions in the square, 

 as in the following examples: — 



F/g 5 Fig 6. 



It will at once be seen that the fir.st of these squares 

 only requires the transposition of 4 and G to become 

 identical with the Natural Square, and that the other 

 by the transposition of 2 and 8 becomes the Magic 

 Square. The only difference in principle between the 

 two squares is that 1 in the first is placed in a corner 



