VOL. LXXV.] PHILOSOPHICAL TRANSACTIONS. 675 



the same spark may be seen divided into a variety of ramifications which shoot 

 into the surrounding air. In this case likewise the fluid is diffused over a large 

 surface, and in proportion to the extent of that surface, so is the faintness of 

 the appearance. A spark, which in the open air cannot exceed J of an inch in 

 diameter, will appear to fill the whole of an exhausted receiver 4 inches wide 

 and 8 inches long. But in the former case it is brilliant, and in the latter it 

 grows fainter and fainter as the size of the receiver increases. To prove the 

 observation, which I think may be justified by the preceding facts, I made the 

 following experiments. 



Exper. 8. To an insulated ball, 4 inches in diameter, I fixed a silver thread, 

 about 4 yards long. This thread, at the end most remote from the ball, was 

 fixed to another insulated substance. I brought the ball within the striking dis- 

 tance of my conductor, and the spark in passing from the conductor to the ball 

 appeared very brilliant ; but the whole length of the silver thread appeared 

 faintly luminous at the same instant. In other words, when the spark was con- 

 fined within the dimensions of a sphere J- of an inch in diameter, it was bright, 

 but when diffused over the surface of air which received it from the thread, its 

 light became so faint as to be seen only in a dark room. If I lessened the sur- 

 face of air which received the spark by shortening the thread, I never failed to 

 increase the brightness of the appearance. 



Exper. Q. To prove that the faintness of the electric light in vacuo depends 

 on the enlarged dimensions of the space through which it is diffused, we have 

 nothing more to do than to introduce 2 pointed wires into the vacuum, so that 

 the fluid may pass from the point of the one to the point of the other, when 

 the distance between them is not more than -^ of an inch. In this case we shall 

 find a brilliancy as great as in the open air. 



Exper. 10. Into a Torricellian vacuum, 36 inches in length, I conveyed as 

 much air as would have filled 2 inches only of the exhausted tube, if it were 

 inverted in water. This quantity of air afforded resistance enough to condense 

 the fluid as it passed through the tube into a spark 38 inches in length. The 

 brilliancy of the spark in condensed air, in water, and in all substances through 

 which it passes with difficulty, depends on principles similar to those which 

 account for the preceding facts. 



5. In the appearances of electricity, as well as in those of burning bodies, 

 there are cases in which all the rays of light do not escape; and that the most 

 refrano-ible rays are those which escape first or most easily. The electrical brush 

 is always of a purple or bluish hue. If you convey a spark through a Torricel- 

 lian vacuum, made* without boiling the mercury in the tube, the brush will 



* If the Torricellian vacuum be made with mercury perfectly purged of air, it becomes a perfect 

 non-conductor. This I believe will be proved decisively by some experiments which I hope will be 

 soon communicated to the u. s. Dr. Price. — Orig. 



4 R 2 



