INVESTIGATION OF ELECTRIC FISHES -- FINAL REPORT - PHASE I

• Document Type: SPECIALCOLLECTION
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): 00173967
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 35
• Document Creation Date: January 21, 2025
• Document Release Date: January 15, 1983
• Publication Date: January 28, 1974

INVESTIGATION OF ELECTRIC FISHES
FINAL REPORT - PHASE I
by
Prepared under:
Contract
28 January 1974
SUMMARY
The African fresh water weakly electric fish Gnathonemus
petersii has been investigated. The study has been directed
toward the intradermal sensory system with emphasis on
the electroreceptors. Three types of electroreceptors have
been identified. The autorhymic activity of these electrore-
ceptors has been recorded. The variation of the electric
signal of the electric organ have been recorded for three
specimens of Gnathonemus petersii as a rest activity and
maximum signal rate when a metallic object has been placed
near the fish. The number and density of different kinds of
electroreceptors in the dermis have been counted and plotted
against rate change and their sensitivity to a metallic object.
The preparation of the large tank experiments have been re-
ported and the newly developed instrumentation is mentioned.
CONTENTS
1.
INTRODUCTION
1
2.
METHODS AND INSTRUMENTATION
2
3.
RESULTS
8
4.
LARGE WATER TANK PREPARATION FOR
EXPERIMENTS
26
111
FIGURES
Figure
1
2
4
5
6
8
9
10
11
Page
African fresh water weakly electric fish
Gnathonemus peters ii 3
Electric fish Gnathonemus petersii in a lucite
restraining tray provided with stainless steel
electrodes 3
Microelectrode amplifier 5
Microelectrode amplifier
shielding tube
Microelectrode amplifier
shielding tube, face side
Microelectrode amplifier
shielding tube, back side
Microelectrode amplifier
form and gain (see Table
and support in its
Microelectrode amplifier
form and gain (see Table
Microelectrode amplifier
form and gain (see Table
Microelectrode amplifier
form and gain (see Table
Microelectrode amplifier
form and gain (see Table
iv
ready to be put in the
6
6
7
input and output wave-
2) 11
input and output wave-
2) 11
input and output wave-
2) 12
input and output wave-
2) 12
input and output wave-
2) 13
ready to be put in the
gure
12 Microelectrode amplifier input and output wave-
form and gain (see Table 2) 13
13 Microelectrode amplifier input and output wave-
form and gain (see Table 2) 14
14 Microelectrode amplifier input and output wave-
form and gain (see Table 2) 14
15 Microelectrode amplifier input and output wave-
form and gain (see Table 2) 15
16 Microelectrode amplifier input and output wave-
form and gain (see Table 2) 15
17 Microelectrode amplifier input and output wave-
form and gain (see Table 2) 16
18 Microelectrode amplifier input and output wave-
form and gain 16
19 Tuberous organ (electroreceptor) of Gnathonemus
petersii 18
20 The electric sensory fields of Gnathonemus
petersii 20
21 Limits of the electroreceptors sensory fields of
Gnathonemus petersii - 20
22 Different types of mormyromasts: a. tuberous
organ, b. A-mormyromast, c. B-mormyromast
(top and cut view) 21
23 The lateral line nerves of the electric fish
Gnathonemus petersii 22
24 Comparison between sensitivity and density of the
electroreceptors of Gnathonemus petersii in the
epidermis 24
Figure Page
25 Autorhytmic activity of the electroreceptors of
Gnathonemus petersii: a. 500 Hz calibration
signal, b. electroreceptors near the chin, c.
electroreceptors near the eye 24
26 Electric activity from the nervus lateral anterior
innervating receptor near the proboscis of a mech-
anical displacement on the chin of Gnathonemus
petersii when the proboscis has been moved upwards:
a. time marks = 50 Hz, b, electric activity in the
nerve, c. movement of the chin proboscis
25
27 12 ft. diameter fiberglass tank provided with heat-
ing, filtering, countercurrent aeration and double
rails for electrode support 27
28 Heating tank provided with automatic control of
temperature to 0.01�C 27
29 Differential amplifier hanging over the water tank. 28
30 Close look at the differential amplifier used in con-
junction with the electrodes in the water tank to
record electric activity of electric fishes 28
31 Differential amplifier with remote control. Ampl.
factor = x 4000, noise = 1 microvolt 29
32 Devices for restraining electric fishes in the water
tank 29
vi
1. INTRODUCTION
In our letter report of 24 July 1973 we mentioned our interest in the
mormyrid electric fishes. One of our reasons is that they have an electrical
quantifiable behavioral variable. The rate and amplitude of their electric
signal changes when they are electrically stimulated or discontinuities ap-
pear in their electromagnetic field. In our study of their electroreceptors
we found some mechanoreceptors on the chin having a close morphology to
the Lorenzini ampulla, a multisensory receptor. This fact confirms our
findings with respect to the Lorenzini ampulla functions from our previous
research.
1
2. METHODS AND INSTRUMENTATION
Three specimens of Gnathonemus petersii (Fig. 1) a mormyrid fresh
water, weakly electric fish from Africa, have been used in our experiments.
The fishes were in our laboratory for three months and each one held se-
parately in 15 gallons aquaria. The water pH was around 6.9 and the tem-
perature was 24�C.
A lucite tray with stainless steel pick-up contacts half embedded in
the lucite wall were used to restrain the fishes during experiments for as
the limits of the repetition rate of their electric activity (Fig. 2).
The microelectrode amplifiers previously developed had a drawback,
namely, when changes in the input resistance occurred, the offset potentio-
meter had to be reset. When very sensitive and delicate microelectrode
recordings were made, resetting of the offset potentiometer was sometimes
an impossible job owing to the diameter of the microelectrode of 0. 5 microns.
The microelectrode amplifier is also the direct support for the microelectrode
proper. Unfortunately until recently nothing was available for the building of
a microelectrode amplifier having all the necessary specs: high input im-
pedance with low capacitance, a reasonable amplification factor, a low output
impedance, very low noise factor and an insensitivity to change in input re-
sistance from 50 ohms to 1 meghom. It also had to use little power and have
a frequency range from DC to 50,000 repetition rate square wave or from
DC to 1 MHz, with an amplification factor flat within maximum 3 dB.
We finally succeeded in building such an amplifier, which could mean
a successful recording under most adverse conditions of small DC or AC
2
40' co,31
�
Fig. 1. African fresh water weakly electric
fish Gnathonemus petersii.
�II � I
11 I 1 j)clIt
kk ti ti �
.
Fig. 2. Electric fish Gnathonemus petersii in a lucite
restraining tray provided with stainless steel
electrodes.
3
signals from the electroreceptors of electric fishes, and which does not
need to reset the offset potentiometer. The amplifiers we used until now
were the best which could be built, but they were far from the capabilities
of the new microelectrode amplifier, which incidentally, could be used in
our simulation of the electric fish capabilities because of its low noise and
very large bandwidth, combined with an insensitivity to change in input im-
pedance. Figure 3 shows the schematic of the new amplifier, and Figs. 4,
5 and 6, are actual photos of the amplifier.
7471,, :NT r �SN er_ MSS.,
IMMO
o
01.47737
o 4-
fill 2 Litnfot Tvr
s1 NoTE; PIN I (Mg. THI Tttb,
1ri-3 M IMELECTRODE AMPLIFIER
Fig. 4. Microelectrode amplifier and support
in its shielding tube.
elk"onal0=411P"Prn
' � �
'
Fig. 5. Microelectrode amplifier ready to be put
in the shielding tube, face side.
Fig. 6. Microelectrode amplifier ready to be put in
the shielding tube, back side.
3. RESULTS
The specimens of the electric fish Gnathonemus petersii were put in
the lucite tray, using their own aquarium water and air was provided through
a special glass tube. The temperature of the water has been recorded. After
a few minutes accommodation to their environment the electrodes correspond-
ing to the head and tail of the fish, were connected to an amplifier, to the
oscilloscope and to a counter. The rest activity has been read on the counter.
Then a carbon steel rod (diam Pe- 3 mm) has been immersed in the tray in the
proximity of the fish. The repetition rate increased significantly and a read-
ing of the counter has been made. Figure 2 shows the fish in the tray. Table
1 shows the repetition rates of the signals. There is a ratio that could go to
1:23 (Fish No. 3) between the minimum and the maximum rate of the signal.
Subsequent experiments could show how the repetition rate and amplitude of
the signals are related to the size, composition and proximity of the objects
The microelectrode amplifier has been checked for its frequency and
gain response using a Wavetek waVegenerator, attenuator and a Tektronix
Oscilloscope type No. 555.
The upper trace of the photos shows the input waveform and lower
trace shows the output waveform of the amplifier. Both sine waves and
square waves have been used. Table 2 shows the waveform, amplitude, gain
and input resistance. Six photos were made for 50 ohms input resistance
and six photos were made for 1 megohm input resistance.
8
::',..:3,ffn!"5.--7,..7.71".?7,7.-- �
7,74.
TABLE 1
Rest and Maximum Repetition Rate of the Electric Signal of
Three Specimens of Gnathonemus petersii
Fish
No.
Date
of
Recording
Weight
of Fish
in Grams
Signal
Water
Temp.
in �C
Instrumentation
Rest
Rep. Rate
Max.
Rep. Rate
Amplitude
mV
1
3
4
8/12/73
8/11/73
8/12/73
15
22
15
15
7
15
136
161
135
500
500
500
20
20
20
Amplif. : 100x
Oscil. Tek. 555
Frequ. Counter
for all Record-
ings
TABLE 2
Microelectrode Amplifier: Input and Output
Waveforms and Results
Photo
No.
Sine
*IA.
Sweep
cm
Gain/cm
Input Res.
In
Out
6
7
8
1K
10K
100K
2 msec
. 5 msec
1 mV
1 mV
1 mV
1 V
1 V
1 V
50a
50C4
502
Photo
No.
Square
_r-L.
Sweep
cm
Gain/cm
Input Res.
In
Out
9
10
11
1K
10K
100K
2 msec
. 5 msec
.05 msec
1 mV
1 mV
1 mV
1 V
1 V
1V
50 CI
50C2
50c2
Photo
No.
Square
_r-L._
Sweep
cm
Gain/cm
Input Res.
in
Out
12
13
14
1K
10K
50K
2 msec
. 5 msec
. 1 msec
1 mV
1 mV
1 mV
1 v
1 V
1 V
1 meghom
1 meghom
1 meghom
Photo
No.
Sine
A
Sweep
cm
Gain/cm
Input Res.
In
Out
15
16
17
1K
10K
50K
2 msec
. 5 msec
. 1 msec
1 mV
1 mV
1 mV
1 V
1 V
1 V
1 meghom
1 meghom
1 meghom
10
Fig. 7. Microelectrode amplifier input and output
waveform and gain (see Table 2).
Fig. 8. Microelectrode amplifier input and output
waveform and gain (see Table 2).
11
-4,���� -���-��-�e�Sr�-��� -��{7.��������}�T�����������,,,,,,,������������-rv�-,_�������.,/*�����,���-�,����*,��� ��� � � .�-���., ���;:��������
Fig. 9. Microelectrode amplifier input and output
waveform and gain (see Table 2).
Fig. 10. Microelectrode amplifier imput and output
waveform and gain (see Table 2).
12
Fig. 11. Mieroeleetrode amplifier input and output
waveform and gain (see Table 2).
Fig. 12. Microeleetrode amplifier input and output
waveform and gain (see Table 2).
13
- Fr.'" 7.7
Fig. 13. Microelectrode amplifier input and output
waveform and gain (see Table 2).
Fig. 14. Microelectrode amplifier input and output
waveform and gain (see Table 2).
14
Fig. 15. Microelectrode amplifier input and output
waveform and gain (see Table 2).
Fig. 16. Microelectrode amplifier input and output
waveform and gain (see Table 2).
15 .
�
Fig. 17. Microelectrode amplifier input and output
waveform and gain (see Table 2).
Fig. 18. Microelectrode amplifier input and
output waveform and gain.
16
����+.7,,,,����������,,�-�arn-�-���,���-_ MeV, -������, ��������517r., .r�Tt. �
For electric receptors we found two types of mormyromasts (A and
B) and one type of tuberous organ. They are confined to well defined
areas of the epidermis. The epidermis of these regions has a particular
structure, which is developed in the Gymnotoides in a similar way. Its es-
sential components are columns of very thin, flat hexagonal cells 60 Am in
diameter, invariable in all species and body sizes.
The height of the columns depends on location, but increases with
body length. The mormyromasts are not covered by the hexagonal cells,
but by small polyhedric cells which are arranged in a circular pattern.
The A-type mormyromasts possess an opening toward the surface and
are evenly distributed with a relatively wide space between them. The B-
type mormyromasts have no opening to the surface, and are more numerous
than the A-type and are also evenly distributed.
The tuberous organs lack an open connection to the surface and form
distinct patterns (Fig. 19). They can be classified according to the number
of their giant sensory cells (1 to 10). All mormyromasts and tuberous
organs are innervated by lateral line nerves. Only the tip of the chin with
its Lorenzini ampullae is innervated by the Nerve trigeminus.
Each mormyromast is enclosed by a loop of capillaries. The common
lateral line system has developed only along the trunk and the tail. In the head
only deep laying canals exist, but without sensory cells.
The tuberous organs are characterized by an autorhytmic activity
yielding a few mV, and with a high repetition frequency exceeding 1 kHz.
The duration of the spikes are approximately 300 �sec. The transmitting
electric organ of Gnathonemus petersii is located in the tailstalk, occupying
2/3's of it and represents approximate 12% of the total length of the fish.
17
Fig. 19. Tuberous organ (electroreceptor) of
Gnathonemus petersii.
18
The repetition rate of the impulses are influenced by light. In day-
light the rest repetition rate is between 7 and 10, in the night it increases
to 15-20. It will also increase considerably in the case of a stimulus affect-
ing the fish.
The EMF of the fish with no load and out of water is between 7 and
17 volts depending on the particular specimen. The internal resistance is
around a few kilo-ohms.
The electroreceptors sensory fields of Gnathonemus petersii can be
clearly visualized if we put the fish in a solution of 10% buffered formaline.
Figure 20 and 21 show the limits of these sensory fields.
There are between 700 and 1000 tuberous organ electroreceptors,
between 800 and 1000 type A mormyromasts electroreceptors and between
2100 and 2300 type B mormyromasts electroreceptors in the skin of an adult
Gnathonemus petersii. The total number of electroreceptors varies between
3600 and 4300. These are distributed on the body as follows: between 42 to
46% on the head on 41 to 44% of the electroreceptor fields; between 30 and 32%
on the dorsal sides on 27 to 30% of the electroreceptor fields; and between
22 and 26% on the ventral sides on 25 to 32% of the electroreceptor fields.
The total area of the electroreceptor fields may occupy between 2000 and
5000 mm2 area for fishes between 90 and 125 mm length. Figure 22 shows
the different types of mormyromast electroreceptors of Gnathonemus petersii.
With the exception of the sensory receptors of the chin which are mech-
anical displacement receptors and are connected to the CNS through the Nervus
trigeminus, the mormyromast electroreceptors are subserved by the lateral
line nerves. Figure 23 shows the main branches of the lateralis nerves
system. All the mormyromasts types (tuberous, A and B) are connected to
nerves forming bundles pertaining to the lateral line system and ending in
the brain.
19
-77 7-7.7
Fig. 20. The electric sensory fields of
Gnathonernus petersii.
- -
,
Fig. 21. Limits of the electroreceptors sensory
fields of Gnathonemus petersii.
20
C.
Fig. 22. Different types of mormyromasts:
a. tuberous organ
b. A-mormyromast
C. B -morm yr omas t
(top and -cut view).
21
t4..7
1A6 L11 RAm4is Doty; LAT, g....04T, SPoiAL (fisiti)
0".tfeit,(4:-:Pi4 .7/1F � .\
�-/- � \
jI
was:
1
torthyitt
/A UCION
01/4/
, -
7.-- �
� - -
�. \' � N\ \
\N-,1(41441 IA � 11 .L.Ar
\ NrM 0r � Ns
LIN/ rf .or
t #t Jekrogy
niTeptrIl
Fig. 23. The lateral line nerves of the electric fish Gnathonemus petersii.
Zit
Ete-frRic
The tuberous organ electroreceptors are autorythmic and the EMF
may reach a few millivolts. The repetition rate varies from 550 to 3900
with the most often encountered repetition rate between 0.95 and 1. 95 kHz.
Figure 24 shows a comparison between sensitivity and density of the
electroreceptors in the epidermis of Gnathonemus petersii and Fig. 25 shows
the autorhytmic activity of the electroreceptors near the chin and near the
eye.
Figure 26 shows the autorhytmic activity of the mechanical displace-
ment sensory organs as an effect of bending the proboscis of the chin.
Experiments in this direction would be continued to record wave-
form and changes in repetition rate as a result of different stimuli.
23
,..
...,./-
:.
� .6�
.i. �
�
% .
---.
.�
� /
T.-
�
,
T-
,
,
if.;
iyii, of
:11440,3
t ,
,
.
[
.
i
, ",e t.titif 4
af rteeito re'
cia!pr.,
�
Jr
.
,
S.
I.0
Is-
PFLATI4 0,:;irir�I
Fig. 24. Comparison between sensitivity and density
of the electroreceptors of Gnathonemus
petersii in the epidermis.
a
b
45mVI
Fig. 25. Autorhytmic activity of the electroreceptors
of Gnathonemus petersii:
a. 500 Hz calibration signal
b. electroreceptors near the chin
C. electroreceptors near the eye
24
a
b
a
b
Fig. 26. Electric activity from the nervus lateral anterior innervating
receptor near the proboscis of a mechanical displacement on
the chin of Gnathonemus peters ii when the proboscis has been
moved upwards.
a. time marks = 50 Hz
b. electric activity in the nerve
c. movement of the chin proboscis
4. LARGE WATER TANK PREPARATION
FOR EXPERIMENTS
The 12 ft. diameter, 4 ft high water tank has been prepared for the
other experiments that will follow for the Phase II investigation (Fig. 26).
Heating the water is done with 2 x 1000 watts heaters controlled by
an "YSI" temperature controller to + .1�C and is normally held at 25�C.
The heaters are in a separate 30 gallon tank and are connected to a relay
switching them on and off and controlled by the temperature controller. Two
9 gallon per minute pumps are pumping in and out the water from the 30
gallon tank from and into the large tank (Fig. 27).
Rails with nylon strings are provided for the silver-silver chloride-
platinized-silver-chlorized electrodes which can be moved from one end to
the other end of the tank (Fig. 28). The electrodes are connected to a re-
mote controlled differential amplifier (ampl. fact. x 4200) suspended over
the, tank and from the amplifier to the differential oscilloscope Tektronix
type 555 (Figs. 29, 30 and 31).
An electric fish can be suspended in one of the restraining devices
shown in Fig. 32. The fish restraining devices are provided with stainless
steel end electrodes which are connected to an audio-amplifier (ampl. fact.
= 300) and to the oscilloscope and displayed on a second beam.
Our preliminary experiments show good promise for recording the
changes in the field of electric fishes produced by them as a result of stimuli
they are presented with.
26
t:Cfh
V 1
�
Pe.
. t
3:1-Alt :77;1
.�
ON) 111.7.4.
ro
I � � -"41.1114/Zipaismil;..,1,,,g_ , � I t
L.
t 4
Fig. 27. 12 ft. diameter fiberglass tank provided
with heating, filtering, countercurrent
aeration and double rails for electrode
support.
�.0
\ /
'4 I s: I - 1 .� , -T1 TT
_
simmul,so mort���� ..c.Nj . 4
/CO . kV it 1. i - rIPP"r"..n.:4;4egrt,r
. , IV ' %I, .:; ',.."....:"..... .s. , fi"t � Ti � :.0 1 �VII
-'
a 41 1...
, j lel. ii�������
111111� i
�
Fig. 28. Heating tank provided with automatic
control of temperature to 0.01�C.
27
�����.e
sort P
.
� � ti.;���
� 7>*1�1:,_
� 4141a1":H- .
t.k
1
(1.1
I ti i Jr.; .
.�
Mal. 41
.1���j���Zt.:=������:.""
I � i:M.
,7.go� iltV4
��
�
� � � ip
�
� ' � �
_
� � it.. k.
Fig. 29. Differential amplifier hanging over the water tank.
't
Fig. 30. Close look at the differential amplifier used in
conjunction with the electrodes in the water tank
to record electric activity of electric fishes.
28
or=u1=5.tews,� =,,,,167,7.7.7�A471tii&t �
; �
Fig. 31. Differential amplifier with remote control.
Ampl. factor = x 4000, noise = 1 microvolt.
Fig. 32. Devices for restraining electric
fishes in the water tank.
29