JxjLY 17, 1908] 



SCIENCE 



93 



this source. At present opinion is divided as 

 to whether the diamond should be considered 

 a constituent of the rock itself as is the mica, 

 garnet, ilmenite, etc., or whether it has crys- 

 tallized at great depths and merely been 

 brought upward by the peridotite. 



The discovery of diamonds at one of the 

 American peridotite localities is causing some 

 speculation as to whether the other two locali- 

 ties where this rock is known to occur and to 

 contain the accessory minerals which so fre- 

 quently accompany that gem, may not also 

 contain the diamond. The remarkable sim- 

 ilarity of the rocks at all three of the Amer- 

 ican localities with those of South Africa, not 

 only in appearance as indicated in the collec- 

 tion mentioned, but also in eruptive char- 

 acter, in inclusions, in structure and in chem- 

 ical composition, as has been frequently noted, 

 makes such a supposition not improbable. 



A catalogue describing the rocks of this 

 collection in detail, together with a bulletin 

 outlining the important facts concerning the 

 Arkansas diamond field has been issued by 

 Commissioner Guy B. Tucker, of the Bureau 

 of Mines, Manufactures and Agriculture. 

 Either of these may be obtained by applying 

 to the Commissioner of Mines, Little Eock, 

 Arkansas. 



Philip F. Schneider 



Little Eock, Aek. 



SPECIAL ARTICLES 



THE EFFECT OF AN ANGLE IN A CONDUCTOR ON 

 SPARK DISCHARGE 



At a recent meeting of the American Phil- 

 osophical Society held in Philadelphia the 

 writer gave the preliminary results of his 

 experimental work to determine the direction 

 of flow of the electrical current in a wire. 

 A large, eight-plate " static " machine, en- 

 closed in a glass case containing also the 

 Leyden jars, was used as a source of elec- 

 tricity. 



The positive and negative discharges are 

 led to two large metal cylinders hung in the 

 air on insulators, and armed with a multitude 

 of pin-points. In this way the positive and 

 negative discharges are not superposed in the 

 same conductor. The discharge was led 



around a sharply made right angle in the 

 wire. This was done by means of a small 

 splinter of bamboo, forming a sharp edge. A 

 photographic plate in a holder of hard rubber 

 was placed under the angle so that the dis- 

 charge could be sent downward to the angle 

 and then led horizontally away over the plate 

 holder. 



By reversing distant connections made by 

 two small wires, the discharge could also be 

 sent around the angle in the opposite direc- 

 tion. 



The plate holder rests on a large sheet of 

 plate glass forming a table top. Below this 

 sheet of glass is a plate of metal connected 

 to the water pipe. Its distance from the 

 photographic plate may be varied. In all of 

 the work done on the negative discharge this 

 plate was not needed. Its use does not 

 change the nature of the result. 



It is found that when the negative dis- 

 charge plunges down to the angle the elec- 

 trical particles pass on into the air and 

 through the rubber cover, whose thickness is 

 three sixteenths of an inch, to the photo- 

 graphic film. This is shown by the character 

 of the image formed on developing the plate. 

 If the action is too strong, the electrical 

 stresses produce branching images due to in- 

 cipient breaking down of the film. By di- 

 minishing the spark length or by removing the 

 angle further from the film, the image be- 

 comes a round spot just under the wire carry- 

 ing the downward discharge. This image is 

 apparently of the same character as is pro- 

 duced by X-rays or radio-action. 



When the discharge is reversed, so that the 

 negative particles pass across the plate to the 

 angle and then upwards, the spot is much 

 feebler, or does not form at all. The result 

 depends somewhat on the extent to which the 

 discharge is an oscillatory one. Oscillations 

 are to be prevented. 



When the positive discharge is sent around 

 the angle, no such effect is produced. This is 

 the case even when the grounded plate is 

 brought as near as is possible without in- 

 ducing spark discharges over the plate-holder. 

 In such cases, however, the plate may be 

 fogged by negative action from below, and 



