OF ARTS AND SCIENCES. 379 



portioi'is of current to wliicli they are severally due. Thus the flame - 

 spark possesses great heatmg power, while the thread-spark is almost 

 destitute of it ; hence the facility with which paper and other combus- 

 tibles are ignited, when placed in a short spark, and the entire failure 

 of this effect when they are made to intercept a long one. 



The fact of the spreading out of the flame-spark under the impulse 

 of a transverse blast of air, while the luminous thread experiences no 

 change, would naturally suggest the comparatively sloiv motion of the 

 former ; and accordingly Robinson of Armagh, by applying the test of 

 "the revolving mirror, has proved that the velocity of the flame-spark is 

 very much less than that of the thread-spark. 



Recently, Du Moncel and Perrot have made numerous experiments 

 on the two kinds of sparks, and the latter, availing himself of the effect 

 of a current of air on the flame-spark, has succeeded in separating the 

 two so completely as to be able to mark more certainly their respective 

 peculiarities. These observations have shown that the current belong- 

 ing to the flame-spark is endowed with decided magnetic and chemical, 

 as well as heating powers, while that of the other is as little efficient 

 in these respects as that of common frictional electricity. In view of 

 these various characteristics, we are entitled to conclude that the flame- 

 spark is due to a current of low tension, like that of an ordinary voltaic 

 arrangement, and the thread-spark to one of high tension, comparable 

 to the discharge of an electrical machine. Whether these currents are 

 to be regarded as simultaneous or alternating in their transit through 

 the coil, remains to be determined. 



The very different character of the discharge through the air when 

 the terminal wires are near each other, and when far apart, is no doubt 

 dependent on the different degrees of resistance which the interposed 

 air presents to the passage of the current. When the terminals are at 

 a short distance asunder, the induced current, having comparatively 

 small resistance to overcome, begins to pass across before its tension 

 has been much exalted, and thus discharges itself continuously during 

 a short time and at a low tension. But when the terminals are widely 

 separated, the current is at first unable to make the transit, and has to 

 accumulate a vei;y high tension at the terminals in order to overcome 

 the greatly increased resistance ; "and when at length it forces a pas- 

 sage, it does so, as might be expected, with a correspondingly greater 

 velocity, and completes the discharge more nearly in an instant. Thus, 



