ELECTRIC LIGHTS. 167 



the lens may be sufficient. If it is not, a screen of 

 paste-board or something similar may be extemporized. 



FLOATING MAGNETS (MAYER'S EXPERIMENT). 



Magnetize six or eight cambric needles so that their 

 points will all have similar poles. Thrust these needles 

 through small vial corks so that when placed in a dish 

 of water they will float with similar poles up. Thus 

 placed they will repel each other and move as far 

 apart as possible. Bring a small bar magnet over 

 them so that the adjacent pole will be the opposite of 

 that of the upper ends of the needles. The needles 

 will be attracted by it and approach it, but repelling 

 each other they will arrange themselves in certain 

 symmetrical order, which will depend upon the number 

 of the floating magnets. If there be but three of them 

 they will assume a triangular form. If there be four, a 

 square. If five or six, there will be two or three posi- 

 tions of stability. To project these motions and forms 

 it will be necessary to have a glass tank similar to the 

 one described on page 47 for cohesion experiments 

 with the vertical attachment to the lantern. The tank 

 must, of course, be deep enough to allow the needles to 

 float freely about. The needles may be short an 

 inch long. If the corks be half an inch in diameter 

 and quarter of an inch thick, they will float in three- 

 quarters of an inch of water without danger of over- 

 turning. The controlling magnet need not be a heavy 

 one. One made of a stout knitting-needle will answer, 

 and it will be best for projecting purposes if the con- 

 trolling pole be bent at a right angle for two inches of 

 its length. This will allow proper movement of it in 

 the field without obscuring the field by large shadows. 

 (See Mayer's experiments, Amer. Jour, of Science 

 1878). 



