Diagrams 3. and 4. illustrate similar plans of connection. In the former the battery B is in the main circuit, in the latter in a shunt to device a.
In Figs. 5. and 6. other modified connections are shown. In 5. an auxiliary secondary S1 with battery and relay is connected around device a which is excited by main secondary S.
In 6. a similar connection is used with a buzzer b to excite device a through secondary S.
Figs. 7., 8. and 9. again illustrate other arrangements in which two
batteries were employed, one generally to secure initial excitation.
By changing connection to the one indicated in the second diagram the condenser was stronger and more effectively charged upon the falling of the resistance of either of the devices a a1.
The conclusion arrived at from many experiments with two devices, which seem to indicate that two such sensitive devices are better than a single one as regards sensitiveness, was that the devices should be arranged as in Sketch 3 so that a change in one will produce a change in the other which in return should react upon the first and so on. This general scheme is to be further considered.
arrangements of apparatus with open secondary for exciting sensitive
This is a modified arrangement (5.) there being only one condenser and besides a very high self-induction around device a to provide for initial excitation when device a is originally of practically infinite resistance. The relay R may be placed around device a instead of self-induction l.
In the diagram below (6.) is shown a manner of connecting apparatus to a circuit Lx C, which is adjusted to be in synchronism with the primary vibrations of the oscillator and excites device a.
Again in Diagram 7. a special synchronized circuit is done away with, the secondary itself being adjusted to the primary vibrations. The plan adopted in New York apparatus of winding secondary and primary on a large drum (8.) serving at the same time as table is best. Tuning is easy, apparatus cheap, a large amount of copper may .be easily placed in the synchronized circuit.
Now resonance of present secondary 5280 feet was obtained with total capacity of 6 tanks instead of 4. Reducing the figures for length we should have for smaller length a capacity larger in proportion , or 1.11 that is instead of 4 tanks we should have had 4.44 for 5280 feet. The required length for 6 tanks capacity. This length would be
With this length the oscillator will require same capacity as extra
coil and good results may be expected.
Various advantages are thereby secured chief of which: double break number, smaller resistance in gaps and increased capacity of W.T. for charging condensers. Connections may be as illustrated in I and II.
This connection shows well with small oscillators provided the short
circuit of secondary of supply transformer avoided.
Here the short-circuit is avoided by the use of self-inductions Ly
and L^ which must be well insulated to stand high tensions. A single
self-induction inserted in the middle wire may also be used with
From reaction of capacity on secondary of W. Transformer it was
probable that more capacity was required, but the transformer was
overloaded when more tanks were joined.
This can be only
due to rapid vibration and suddenness as tension on that box only
15,000/2 volts. There
are some doubts as to the distribution of e.m.f. in condensers in
series when vibration takes place. Strong (probably inductive) drop
on supply circuit (exceptionally so).
Connection was changed to that indicated in sketch for the purpose
of avoiding effect of short circuit of secondary of W. T. through
primary arc. Absorbed energy was great. Sparks on switch serious.
Lamps would go up very much when arc would break through. But
general results not satisfactory. The condensers directly on W.TS.
take strong current.
Plan illustrated in first diagram was adopted to obtain double number of breaks with same disk and securing other advantages. Also to better utilize W. Transformer. It was found that when one side on it worked remarkably well, sparks about 4 feet. The tension on each half of transformer being 11,000 volts approx. when both parts on interaction hurtful. The chief drawback being short circuiting of secondary.
The arc was snappy and loud indicating short circuit and
rapid vibration through wire W. The secondary discharge was thick
but spark not long about 3 feet. All tanks were in on
either side and the transformer charged them full when separate.
When both parts on evidently the secondary of W. T. was overloaded.
In Diagram 1. first experimented with, a disadvantage was found to exist: namely, the receiver R was operated through the break device d. This inconvenience was done away with in arrangement illustrated in Fig. 2. which allowed more sensitive-adjustment of Relay R and the apparatus worked better.
Capacity of condenser was varied up to
20 mfd. with changing success. Best results seemingly with small
capacities up to 1 mfd. Secondary about 4,000 turns, primary Lamp
cord No. 10 turns 28. The apparatus responded freely to small pocket
coil at a distance of several feet with no capacity attached and no
adjusted circuit. Consequently will go at great distance.
adjusting speed of rotation of sensitive device the inconvenience is
Gaps were 1 1/16 on each side plus gaps in box 2 turns. Tuning remarkably exact, 1/8 turn of self-ind. box reducing the effect very much. When exactly 4 turns in box, sometimes streamer 8 foot long would shoot out from a defective spot on wire. The ball on top reduced streamer capacity and prevented streamers from coming out all along the top turn as usual. The spark gaps work extremely well, loud explosive character indicating good vibration.
Such sparks are always noted
when secondary well tuned. The system worked economically, the lamps
in supply circuit not falling at all. The earth connection now was
taken off and oscillator of same period (the secondary 29 turns connected). Both had now
same period, the secondary and the extra coil. On first throw of
switch a spark darted to roof above from the ball and the cord
caught fire. Fortunately, it was extinguished before doing damage.
The sparks jumped from the upper turn of
the coil to the Earthed ball and sparks of seven feet were easily
obtained. It was evident that the distance could be much increased
but this was deemed hazardous. As it was the sparks of seven feet
were probably the longest obtained from such large balls or surfaces
of such small curvature.
As tiftie pressed, further experiments with the view of ascertaining the fundamental note were postponed and the first connection with one primary turn again made. Both balls were now connected in multiple to top of the coil and to the upper rod of a spark gap, the lower rod being earthed. There was no danger of setting fire in this way. The display was remarkably noisy. The sparks were up to 14 feet long, snapping quick,-explosive and very white.
Sometimes streamers would shoot out fully 11 feet. Often
several sparks at once. No particular direction in striking. The
capacity in the primary circuit was varied up to 8 tanks on each
side. Always striking effects. The ground wire had no capacity and
no sparks were seen on arresters but before, with only one ball and
no streamers, sparks of 5/16" were drawn from water pipe in distant
Very powerful streamers were produced sometimes extending the full distance between the wire circles, but still they showed tendency upward in spite of presence of ground circle. Often sparks would pass in curved paths between the two circles. During the display no sparks on arresters, small sparks in adjacent room from water pipe. Capacity on each side from 5 2/37 tanks.
streamers with former value. The circle of 8 feet diam. was then
taken down and another one about 10 feet placed on top of coil.
Streamers now showed some tendency to pass to grounded circle.
Sparks to the latter more frequent and brilliant. No play on
arresters and small sparks in adjacent room as before.
Experiments with oscillator secondary 29 turns and
extra coil last described continued.
In the first trials connections were made as in sketch. The spark
gap between wire circle on top of extra coil and ball supported was
Otherwise it was
surmised that the vibration of the secondary itself with the extra
coil might be responsible for most of the e.m.f. generated between
the ground and line. Singularly, despite this strong effect as
evident from arresters but very small sparks were drawn from water
pipe in adjacent room, this seemingly indicating that in this
experiment the earth acted as a nodal region.
As the extra coil had now only a small initial pressure the
e.m.f. obtainable in the spark gap was much smaller and the gap was
to 4 feet at which distance sparks readily jumped. The play on
arresters though weaker
If a nodal region then the e.m.f. set up at the small distance of 60 feet separating the point of attachment and the ground of lightning arresters was only a small part of the total e.m.f. But if a crest then the e.m.f. set up and causing sparks was nearly the total e.m.f. produced by the apparatus. If a nodal region near the point of attachment of the secondary, then at a distance of about 4000 feet there must be a strong effect, but if a crest, then at that distance there would be no effect. This is to be decided by further observations.
The connection was now changed to that indicated. It was thought that both vibrations would cooperate and produce a stronger effect, but it was at once evident that so long as streamers (which were about 10 feet) formed on top of the extra coil the effect must be smaller, since all energy came from the secondary and the streamers caused loss. A condenser ought to be used instead of a gap to make such an arrangement most economical.
Although owing to nodal point the length of spark in adjustable gap was small, the display was strong on arresters, but not nearly as strong as when the extra coil was entirely left off. In the latter case the action was very rigorous so that often flames would form on arresters showing short circuit of dynamo. Also the other choking coil would break through. Evidently then the extra coil did not in this instance prove useful in intensifying vibration contemplated.
Experiments continued: extra coil was
now lowered 2 feet nearer to ground, distance now being about 4 feet
from floor and 5 feet from ground.
The gaps outside 1 1 /4" each approx. Inside
1 1 /22 turns. Streamers produced were still more powerful being
made so owing to approach of secondary. They would dart out to a
distance of 12 feet sometimes.
To test this a coil 50 turns referred to before was inserted in the ground wire of the lightning arresters. It was expected that it would weaken discharge across, but did not probably because the current was small and the choking action likewise for this reason.
To see whether there is some current passing through the earth wire
to the line, another coil was placed in inductive relation to the
ground wire coil and strong sparks 3/8" were obtained on former.
Sparks, lively 3/8" approx., were also obtained from coil P. Note:
Sparks to ball sometimes, at other times streamers would dart past
the ball. The streamers horizontal when sudden, when switch was held
longer they would waver. In last experiments only half of wire
circle on top of spool was used.
This is illustrated in a general way in the
little diagram below. The battery strains the device a through the
condensers C Ct but when, upon the device a becoming excited, the
condensers are suddenly charged the current impulse caused by the
charging automatically stops. It is then necessary to reverse the
mains, or discharge the condensers to make the apparatus ready for a
second operation. This plan allows use of very high pressure on the
sensitive device which should be of great resistance.
Abundant sparks and streamers were produced. The play on arresters was also observed at each throw of the switch. The rain and lightning were just beginning. Magnificent intense white light witnessed below Pike's Peak, something very unusual. It resembled a white hot silver furnace. The lightning on the mountains was very frequent and the discharges of unusual brilliancy. Twice a curious phenomenon was noted. Lightning striking in one part of the mountains from cloud to earth, there was seen in another part a few miles away from a high peak a lightning discharge which to all appearances came from the peak to the cloud.
The discharge was much thicker at the root and branched out towards the sky spattering itself in many branches and disappearing in fine streams. The astonishing phenomenon was witnessed a second time and subsequently, though there was much uncertainty about the direction in the latter cases; a few times a similar discharge took place from other peaks. Is it possible for a discharge to go from Earth to cloud? As far as the visual impression is concerned there can be no doubt.
The discharge in all cases followed a preceding lightning discharge in another region, and apparently from cloud to earth.
Perhaps it can be the effect of an intense vibration started by the
first discharge which results in another discharge towards an
oppositely charged cloud. The clouds were unusual in configuration
and grouping. A large portion of the sky was quite clear. The wind
at times was very strong. An instrument by its constant play
indicated strong electrical disturbances through the earth, even
when there was no display of lightning as far as could be seen or heard.
This gave a smaller streamer surface and consequently longer streamers. The display was fine. In order to see whether the upper half circle of bare wire was responsible for sparks on arresters the wire was replaced but still no result. Then it was thought that other causes for the sparks not appearing were responsible and everything that could have the slightest bearing upon this was investigated. Still nothing was arrived at. The sparks did not appear no matter what change was made in the adjustment of the vibratory circuits. What could be the cause?
The only explanation at present is that the roof
was rendered slightly conducting (although there was little rain in
this locality) and that this produced the change. Important to find
out. Observation: The lightning lighted two houses about two miles
Accordingly, the connections were made as in sketch, the method of magnifying by oscillating transformer being made use of to increase sensitiveness. As far as practicable all connections and parts of instrument were used.
The connections of primary circuit including break
same, only a battery B. (14 cells dry O.K.) and sensitive device a
being inserted instead of a generator. In the high tension secondary
were connected a receiver R (relay), telephone T, battery B, and
another similar sensitive device a1. The motor was driven from a
small direct current generator which in turn was driven by the
alternate current motor usually employed to drive the break disk of
the large oscillator. This apparatus was extremely effective, merely
the addition of small capacity on a was sufficient to make the
receiver respond. Evidently this effectiveness is due to the
efficiency of the oscillating transformer and excellent working and
high frequency of the mercury break.
But on the first throw of the switch the streamers and sparks darted against the netting a distance of about 12 feet and sparks were seen to go from netting on to the wooden structure of the roof. It was advisable to stop work and the roof was removed. Now the ball on top of the extra coil was connected to the latter by a wire No. 10,40 feet long, very heavily insulated with tape over the rubber covering.
One turn on the outside and nearly another complete one in the inside were made and the end of the wire connected to the ball. The latter could not be lifted up and the experiment was tried with the ball in place. The streamers now appeared on the ball copiously when the current was turned on, their tendency being to go straight up into the air.
longest were only about 4 feet as it was deemed unsafe to strain the
apparatus higher until further provisions for safe working were
made. The lightning arresters were observed but no sparking. This
showed that the absence of sparks was not due to rain or moisture as
was concluded yesterday, since the weather was very warm and dry.
The wire was ordinary magnet wire No. 20, 516 turns. The self-induction was approximately calculated from the following data: diameter of drum 25 1 /4" or 64 cm.; length of wound part 20" or 50.8 cm.
The coil was now tuned with oscillator in response to a somewhat
higher note with small capacity on free terminal, the other being
connected to the water pipe. Sparks of 3/4" were obtained while from
the water pipe alone a very minute spark, scarcely perceptible,
could be obtained. Induction from primary being carefully
eliminated, the sparks were still 3/8" long and white.
Still more curious it was to
see a moth at a distance of fully eleven feet from ball, near to the
wooden frame fall straight down as the switch was thrown in. The
strongly electrified ball evidently exercises a strong attraction on
a small insect which is drawn towards it every time the ball is
electrified. This was repeatedly tried.
The action of the wind suggests the idea of preventing the formation of wasteful streamers by a current of air.