Mr. C. Tomlinson's Experiments on the Electrical Fly, 213 



metallic contact, a disc of tinfoil was attached to the bottom of 

 the vessel (see fig. 3). The fly moved very easily on its point 

 within the vessel. 



On pouring in enough turpentine to cover the fly completely, 

 and holding an uninsulated point over the vessel, the liquid 

 was thrown into great commotion, welling up the sides of the 

 vessel and spitting out numerous globules. But the most re- 

 markable effect was on the fly : it rotated rapidly and uneasily, 

 frequently rising and falling on its support ; but its motion was 

 not, as in air, with the points backwards ; the points were now 

 moving forwards. I next connected the chain with the point 

 over the vessel, insulating it on a cake of resin, and connected 

 the finger-glass with the earth (see fig. 4). There was still much 

 mechanical disturbance, but the fly constantly rotated with the 

 points forwards. 



When benzole was substituted for turpentine, there was a 

 greater mechanical disturbance than before, but the fly rotated 

 with the points forwards. On trying the action of a knob instead 

 of a point over the liquid, a spark passed and kindled it. The 

 flame was immediately extinguished by placing a copper disc 

 over the vessel, when, on working the machine and holding a 

 point over the vessel as before, the fly rotated with the points 

 backwards as in air. 



In' paraffin oil the fly rotated at first with the points back- 

 wards as in air. On distilling the oil and repeating the experi- 

 ment, the fly rotated with the points forwards. I obtained this 

 result with several kinds of paraffin oil ; and on purifying them, 

 and two or three specimens of turpentine by distillation, it 

 seemed as if the fly rotated in one direction in the crude oil, and 

 in another direction in the distillate. By multiplying the expe- 

 riments from day to day, contradictions crept in which showed 

 that I was not on the right track. It then occurred to me that 

 the finger-glass, though 4| inches in diameter, did not present 

 a sufficient surface of liquid to enable me to study the effects 

 that were going on in this miniature maelstrom, where the con- 

 fusion was so great that nothing could be distinctly seen, and 

 the mechanical action was so violent as often to throw the fly 

 from off its support. 



I therefore proceeded to examine the action of electricity on a 

 liquid dielectric contained in a shallow glass dish, 6 inches in 

 internal diameter, and less than an inch in depth in the centre. 

 Paraffin oil was poured into this, and the wires of a universal 

 discharger were made to dip into it at opposite points near the 

 rim, one wire being connected with the prime conductor, and 

 the other passing to the earth. The moment the electrical 

 machine was worked, a wave started from each wire, showing 



