180 The N.Z. Journal of Science and Technology. [Nov. 
An interesting test (fig. 3) which more nearly approximates to actual 
conditions was carried out by putting the insulator under a pressure of 
40,000 volts into the humidity-box and then raising the humidity to a 
high degree. By a suitable arrangement of switches, megger-readings could 
be quickly taken after switching oft' the pressure. The curves show that 
the surface resistance fell rapidly to 300 megohms and remained stable 
at that value. Voltage was then removed, and the resistance dropped 
rapidly to about 50 megohms, and showed a tendency to become steady at 
about this value. Pressure was then again applied, and gradually raised 
to 40,000 volts. Under these circumstances the surface resistance again 
increased to the 300-megohm value, at which it remained stable again. 
This test represents to some extent the conditions which may exist on a 
line which is taken out of service during a mist, as, for instance, by the 
breaking-down of a line. Under such circumstances it has often been 
found difficult to place the line into service again, repeated breakdown 
{a.) ( b.) 
Fig. 2.—Humidity tests of 66,000-volt insulator— (a) surfaces clean; ( b) surfaces 
dusty. No. 1 curve shows resistance head to pin; No. 2 curve shows 
resistance head to first joint.p 
resulting on each attempt. Several times a line, previously quite sound, 
has been put under pressure during a mist, and in almost every such 
case it has broken down. In the early stages of operation, when only 
one line was required in service, it was the practice to put pressure on the 
spare line at midnight daily. A number of breakdowns of this line occurred 
within a few minutes after full pressure had been reached, voltage being 
run up in a quarter to half an hour by a spare generator. The hour of 
running up pressure on the line was altered to early evening, when the 
humidity of the atmosphere was still low, and since this alteration no 
breakdowns have resulted from this cause, thus indicating definitely that 
the pressure breakdowns were promoted by the condition of the atmosphere. 
It was also found that if pressure were taken oft a line during misty 
conditions and put on again gradually within a few minutes it was not so 
liable to breakdown as if a longer period were allowed. Apparently there 
is a condensation of moisture on the surface which requires an appreciable 
time to take place and is prevented as long as pressure is maintained, but 
takes place gradually as soon as pressure is removed. 
