108 
The N.Z. Journal of Science and Technology. [June 
THE FAILURE OF HIGH-TENSION INSULATORS IN A 
DAMP ATMOSPHERE. 
By Professor E. Marsden and G. F. Ferguson. 
It has been noticed by Birks and Ferguson (“ Notes on High-tension 
Insulators,” this Journal , vol. 3, pp. 177-90, 1920) that breakdowns 
of high-tension insulators are in certain circumstances directly associated 
with atmospheric humidity. If a line which has been dead for some hours 
be charged during the night, when the humidity is high, a breakdown often 
occurs within a few minutes of charging ; while if the line be charged 
during the day, when the humidity is less, breakdowns are very infrequent. 
This effect is so marked that it has become necessary in the case of the 
Lake Coleridge transmission-lines to make a practice of always charging 
the spare line by day and keeping it charged until required. The object of 
the present paper is to indicate a reason for this. 
It is well known that there is a leakage of current from transmission¬ 
lines due to conduction through the ionized air. Owing to the relatively 
small number of ions in free air this leakage from a thick conductor is 
generally small. When surface irregularities or projections from the wire 
occur, however, the loss is serious, owing to the high potential gradient 
near the projection breaking down the air and causing intense ionization, 
which conducts the current. Such losses are well known, and are termed 
“ corona ” losses owing to the bluish brush-discharge or corona which 
accompanies this effect. The potential gradient necessary to break down 
the air in this way is of the order of 75,000 volts per inch. Even with thick 
wires, which give a much smaller potential gradient at their surfaces than 
thin wires, the corona is shown at various points along the wire where, 
as has been stated above, there are surface irregularities or projections. 
Thus in general there will be a considerable degree of ionization in the 
neighbourhood of a charged line where these irregularities are more 
frequent. 
Electric ions have a very interesting physical property of acting as a 
nucleus for the condensation of moisture from a supersaturated atmosphere 
such as occurs at night. This property is best illustrated by placing a 
highly charged needle-point, from which electricity is leaking, next to a fine 
jet of steam. The discharging needle-point alters the character of the jet, 
causing it to condense and thus become opaque. 
Consider a supersaturated vapour, and let us assume that a small drop 
of water of radius r has condensed. The inward pressure on the drop due 
to surface tension is 2 T/r, where T is the surface tension of water = 74 
dynes/c.m. As this pressure is inversely proportional to r it will be very 
great for small values of r, and will, in fact, prevent the very formation 
of the drop unless dust-particles are present to give the water-vapour a 
relatively extended surface to condense on. Suppose our hypothetical 
drop has a charge e however, there will be an outward electrostatic 
pressure = 2 rr a 2 per unit area where <r is the surface density of electrifica¬ 
tion. Since <x - e/4 7 r r 2 it will be seen that when r is small, even for a 
small value of e the outward pressure will be great compared with the 
inward pressure due to surface tension. This is because r 2 occurs in 
the denominator and makes the expression very large when is small. 
