THE PRESSURE UPON THE POLES OF THE ELECTRIC ARC. 
125 
(l) Using alternating current the arrangement shown in fig. 18 was adopted. No 
current passed through any of the movable parts of the apparatus. An arc was 
struck between two vertical carbons below the end of the suspended pole. The 
torsion head showed a reduction of pressure amounting to 0’4 dyne. The effect 
persisted after the circuit was broken, and until the poles cooled down. 
This quantity is to be added to the observed pressure upon the poles to give the 
true reaction. 
( 2 ) The apparatus was arranged as in fig. 19 for the double arc method. No 
current flowed through the main apparatus, but an arc was started 1'5 cm. below the 
arc gap 'between two carbon rods placed parallel to the suspended carbon; with a 
direct current of 8 amperes a negative pressure of — 0’32 dynes was recorded. It is 
probable that the convexion pressure is actually greater than this, because the arc 
was necessarily some distance below the suspended poles. 
( 3 ) In an experiment with 8 amperes direct current (double-arc method) the 
suspended pole was shaped as shown in fig. 20, with the object of preventing the end 
from growing pointed, but this was partially defeated by the deposition of a small 
point of carbon upon its centre, like a boss upon a shield. The disc on the right-hand 
side was too large (3 cm. in diameter) to be heated over its whole area by the arc 
above it, hence the convexion currents due to the two sides did not balance. For a 
T 
VOL. CCXX.-A. 
