370 The Rev. T. R. Robinson dn the Stratification 



these sparks may be found relevant to the present question. If the 

 terminals be connected with a spark-micrometer opened to about 

 twice the striking distance, when the machine actSj a lucid star 

 will appear on one or both points. Bringing them gradually 

 closer, a small brush, exactly like that produced by a point on a 

 prime conductor, is seen at the positive point : as the distance is 

 lessened, the filaments of this brush extend, and at last curve 

 towards the negative with a sputtering sound. Still nearer, and 

 sparks strike across with an intense light and sharp snap which 

 cannot possibly be mistaken for the preceding form of discharge : 

 they are zigzag ; and, when the excitation is powerful, two-thirds 

 of their length at the positive end is often red, the rest bluish 

 white. If the distance be less, the spark has that strange yel- 

 low envelope which, as Du Moncel has shown, can be blown 

 aside like a flame, but which is certainly not a second discharge*. 

 And at very short distances there is sometimes a sheaf of curved 

 sparks between the points. If one terminal be connected with 

 the gas-pipes of the house, and the arm of the micrometer which 

 had been joined to it be connected with the floor (so that the 

 circuit includes a very great resistance), we obtain what is 

 called the static discharge, which is of the same character 

 as the inductive one discovered by Mr. Gassiot, and, like it, 

 can be distinguished from the dynamic one by the magnet 

 when passing through a vacuum and the revolving mirror 

 in air : it is about half the length of the other. The vacuum 

 which I included in the circuit was an " electric egg," 9' high 

 and 6' diameter; on its wires were cemented glass tubes with 

 Wollaston's points of platinum, -gL diameter, 6' apart. This 

 form of electrodes reduces the conditions of discharge to a more 

 definite state than when they are balls or naked wires ; and the 

 following may be considered its normal character when the egg 

 has been filled with dry hydrogen and exhausted to 0''08. Sup- 

 posing the upper electrode positive, there is at it a brilbant 



whose length is 1'3 times that of those obtained by the spring rheotomes 

 of Hearder or Ladd : Ruhmkorff's vibrating hammer is much feebler. 

 This very energy, however, tests the insulation of the coils severely. I 

 find two precautions necessary in using it: — (1) the mercury must be nega- 

 tive, otherwise the action is almost exjilosive, and the effect only f; (2) the 

 platinum points should work through a diaphragm of thin vulcanized rub- 

 ber, to prevent the blackened alcohol from being splashed about. 



* When such a spark is viewed in a revolving mirror, its thickness is 

 slightly increased, but the envelope is drawn out into a sheet extending 

 many degrees, even when the rotation is comparatively slow, which shows 

 that it lasts much longer than the spark itself. This explains a singular 

 fact ^vhieh has occurred to me occasionally. In a hydrogen vacuum =2'"50, 

 if a small Leyden jar be connected with the machine, accorthng to Mr. 

 Grove's plan, the discharge passes as a splendid scarlet spark; this is 

 sometimes surrounded by a faint elliptic envelope, which continues visible 

 ion 0^*1 after the other has disappeai-ed. 



