184 
The N.Z. Journal of Science and Technology. [Nov, 
all the moisture was dried out of the porcelain by the oven treatment, since 
moisture which takes a year or two to penetrate the porcelain is not 
likely to be driven out in a couple of weeks’ drying. Probably the drying 
only affected the porcelain to a very small depth, but sufficient to show 
up well on the tests provided. 
It is important in connection with the results set out in Table 2 to 
notice the extent to which insulators definitely faulty in one or more shells 
can still maintain full pressure and give continuous service. One very 
striking case occurred at Sockburn, about three miles from the Addington 
(Christchurch) end of the line. A short circuit occurred on the line at 
9.37 a.m., and on testing the line was found good, the short circuit being 
evidently due to a flash-over. Pressure was therefore restored and the 
line put into service again. A few minutes later word was received by 
telephone from Sockburn that an insulator on the line had exploded with 
a great flash. The patrolman got away at once, although very sceptical 
in view of the fact that the line was actually in service again, and to his 
surprise found that the report was correct. Three out of the four shells 
were shattered, but the fourth shell was still intact, and in spite of a light 
drizzle was holding the full pressure by itself. As the load was very 
heavy and the atmospheric conditions unfavourable to taking a line out 
of service, it was decided to postpone the replacement until later in the 
day, and it was actually replaced at 4.45 p.m., the single shell carrying 
full pressure satisfactorily under unfavourable weather conditions for that 
period of seven hours. Tested after removal the single remaining shell— 
the inner—flashed over at 85,000 volts. 
These and similar experiences raise the whole question in the design 
of high-tension insulators as to whether the large quantity of porcelain 
usually employed would be essential if a perfectly vitrified and reliable 
material were available, and whether the quantity of material and number 
of shells is not rather intended to meet the practical certainty in the past 
that a large proportion of the material will deteriorate and prove defective 
after a few years’ service. If this is so, is it not possible, when the manu¬ 
facturers can guarantee a thoroughly vitrified and reliable material, to 
reduce substantially both the size and the number of shells required ? 
That porous porcelain is a serious menace is shown by the fact that 
some insulators tested proved to be quite good conductors. Such insulators 
conducted the full-load current of the 20 K.V.A. testing equipment at a 
voltage too small to be read on the 100,000-volt electrostatic voltmeter. 
The shell under test would crack in a few seconds, and small clouds of steam 
rise from the joints and cracks. Pieces of these shells broken off and tested 
separately exhibited the properties of an electrolytic conductor— i.e., rapid 
increase in current with increase in temperature of the moisture content. 
Insulators which are shattered to pieces in service are no doubt exploded 
by the sudden transfer of the entrapped moisture into steam, due to the 
passage through the insulator of breakdown current. Pieces of such insu¬ 
lators have been projected nearly 100 yards by the violence of the explosion. 
During an experiment on a piece of waterlogged porcelain small fragments 
of the piece under test were thrown off violently by the passage through 
the specimen of about 0-2 ampere. 
A valuable series of experiments on porosity were carried out by 
Dr. Coleridge Farr, of Canterbury College, as described in a paper entitled 
“ The Porosity of Porcelain, with Special Reference to High-pressure Insu¬ 
lators for Electric Transmission-lines,” published in this Journal, July, 
