41 G Prof. Caftan on an Induction Coil of great power. 



from the plate was less than -i inches, and the distance between 

 the circumference of one plate and that of the other was 10 inches. 

 But when a pointed wire projected more than three- fourths of an 

 inch from the middle of a plate connected with the positive end 

 of the coil, and the other plate was connected with the negative 

 end, the sparks passed from the point to the negative plate, and 

 never from the circumference of one to that of the other. 



Sixthly, when a pointed wire projected from the negative plate 

 at about half an inch from its circumference, and 3| inches 

 from the plate, the sparks passed even then far less frequently 

 between the point and the opposite circumference of the positive 

 plate than between the circumference of one and that of the other 

 plate. But no such effect will take place when the pointed wire 

 projects from the positive plate. 



When a pointed wire projected 3| inches from the middle of 

 a wet slate about 9 inches square, or 2| inches from the middle 

 of a wet piece of wood of the same size, connected with the nega- 

 tive end of the coil, and the 12 j-inch plate was connected with 

 the positive end, the sparks passed between the circumference of 

 the plate and the edges of the slate or wood, rather than between 

 the point and plate. When the slate or wood is dry, the sparks 

 passed from the pointed wire to the positive plate. When the 

 wood or slate is merely clamp but not very wet, the sparks from 

 the wire frequently run to the edge of the slate or wood and then 

 pass to the plate. If the wire project from the wood or slate 

 near the edges, the sparks will pass between the point and posi- 

 tive plate, unless when the pointed wire projects to a small dis- 

 tance from the wood or slate. 



A ball 3 inches in diameter connected with the positive end 

 shortens the spark as much as a 12-inch plate. 



When a pointed wire is opposite the edge of a plate connected 

 with the negative end, the spark is longer than between two 

 points, but shorter than between a positive point and the middle 

 of a negative plate. 



When two plates are connected with the ends of the coil and 

 face each other, the spark is reduced from 15 inches to about 11 

 inches. When their edges are opposite to each other, the spark 

 is also shortened. 



I have repeated most of the above-mentioned experiments with 

 a weak battery which gave sparks 7 or 8 inches long, and obtained 

 the same results. When the sparks are about 8 inches, the plate 

 connected with the end plate should not be more than 7 inches 

 in diameter. When I connected the 12|-mch plate, or a 3-inch 

 hollow ball, with the positive end (the longest spark being about 

 8 inches), no spark passed between the plate or ball and a 

 negative point until the point was brought within about 2 



