18S6.] electric discharge in a uniform electric field. 401 



moving in the direction of the lines of force, and we should expect 

 that any change in the shape of the vessel which diminished 

 the average time the molecules were moving in the direction of 

 the lines of force would raise the critical pressure. 



We must remember that the quantity we are comparing the 

 dimensions of the vessel with is not the mean free path of the 

 molecules, which no doubt at the critical pressure is very small 

 compared with the dimensions of the vessel, but the average 

 space taken to reverse the direction of motion, which may be very 

 much greater. For if the force between the molecules be very 

 small, it will require a great many collisions to reverse the 

 direction of motion of the molecules. 



The well-known phenomenon that a thin layer of air is 

 relatively electrically stronger than a thick one, might be explained 

 in the same way as we explained the dependence of the critical 

 pressure upon the shape and size of the containing vessel. For 

 consider the case of two parallel planes, when the distance between 

 the planes is comparable with the average space through which 

 the molecules move continuously in one direction, the constraint 

 caused by the planes will have the effect of diminishing this 

 distance, and therefore of diminishing the chance of the molecule 

 being decomposed by the electric field, it will therefore make the 

 layer stronger than an equally thick layer of gas at the same 

 pressure placed in open space. 



The same explanation will also apply to an experiment of 

 Hittorf (Wiedemann Lehre von der Electricitdt bd. iv. § 417) ; 

 in this a sealed glass flask is furnished with two electrodes 

 made of straight pieces of wire, one being put in the neck, the 

 other in the bulb of the flask. When the one in the bulb is 

 negative it glows all over, but when the one in the neck was 

 negative it only glowed over the tip. In this case the electrode 

 is near the glass sides of the flask, so that the average distance 

 the molecule is moving in the direction of the lines of force is 

 diminished by the glass in those parts of the gas away from the 

 tip more than it is in those near the tip, so that the gas away 

 from the tip is electrically stronger than that close to it. We 

 should therefore expect the discharge to take place at the tip, 

 which is exactly what happens ; on the other hand, when the 

 electrode is in the bulb the glass is a great distance from the 

 electrode, it can not therefore affect the distance a molecule 

 travels before its direction is reversed : the gas is therefore equally 

 strong all over the electrode, and we should therefore expect, as is 

 the case, the electrode to be covered with glow. 



These considerations would seem to have an important ap- 

 plication to the case when there is a contraction in the cross 

 section of the discharge tube at some point. If the tube is 



vol. v. pt. vi. 28 



