Jccoiint of some Electrical Experiments. 163 



tinued to increase to the 200clth explosion when it had produced 

 a projection of 18 linos diameter and three lines high, but no 

 perforation was apparent. This cylinder was sawed through, 

 and the expansion which had taken place in the interior appeared 

 to have been produced by the action of some power determined 

 in every direction from a centre. M. De Nelis appears to con^ 

 sider these experiments as a demonstration of the different mo- 

 lecular attraction of the electric fluid for various metals : but I 

 must confess that to me they seem to present no such evidence, 

 nor do they appear to differ materially from the usual expansive 

 effects of the charge. W'hen electricity passes luminously through 

 any fluid, it invariably produces a sudden and temporary expan- 

 sion thereiji, and the expansion is greater in proportion to the 

 quantity and momentum of the electricity, and to the noncon- 

 ducting power of the Huid through which it passes : hence the 

 spark is brighter and the expansion more considerable in oil, 

 alcohol, or ether, than in water ; more feeble in hot water than 

 in cold, and still less evident in saline fluids and concentrated 

 acids ; but in every case the expansion is greatest when the bulk 

 of the fluid in which it occurs is least, and when the spark by 

 which it is produced passes through the greatest interval. The 

 arrangement employed by M. De Nelis is well calculated to ob- 

 tain these essentials ; for the small leaden wire which he has sub- 

 stituted for the interrupted circuit of former electricians, admits 

 the passage of a much more considerable charge than could be 

 conveyed through the oil or water itself by the same power ; 

 whilst the fluid employed is so arranged as to occupy the least 

 possible space compatible with the luminous passage of the 

 charge through it. 



In the experiments I have made on this subject I have been 

 favoured with the assistance of my friend Andrew Crcsse, Esq.; 

 and it w.is by the aid of his powoiful apparatus and active ex- 

 ertions that 1 was enabled to shorten the time of their per- 

 formance materially. M. De Nelis generally employed a battery 

 of 100 feet of coated surface, which he charged to about 60 of 

 Henley's cjuadrant electrometer ; tliis was most probably equal 

 to 15 grains of Cuthbertson's discharging electrometer, at least 

 it is so when the index of Henley's quadrant is terminated by a 

 pith-ball ; but a cork ball will only indicate about 40 with the same 

 charge. 1 5 grains appeared therefore the most convenient power, 

 and was employed in tliese experiments most frequently. The 

 battery we made use of consists of 50 coated jars, exposing 

 about 75 feet of coated surface ; it was charged by two cyhnder 

 machines, one of them 52 inches circumference, and the other 

 40 inches only. The large machine T constructed in the year 

 1308 : ii is on the whole the bpst 1 have yet seen. The smaller 

 L 2 one 



