LETTER TO/FROM DELETED RE PROPOSAL WITH ATTACHED REPORT TITLED "A PROPOSED THREE-PHASE PROGRAM TO DEFINE HARDWARE ANALOGUES OF ELECTRIC FISH SENSOR SYSTEM

• Document Type: SPECIALCOLLECTION
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): 00176559
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 49
• Document Creation Date: January 21, 2025
• Document Release Date: January 15, 1983
• Publication Date: August 31, 1973

3/ � /5'7tj
Dear
As a result of our discussions on 21 August 1973, I am propos-
ing a three phase program designed to:
(1) Establish the effectiveness of electric fish in detecting
foreign objects at a distance;
(2) Determine the methods used by electric fish for ranging
and location; and
(3) Design equipment systems suitable for replacing electric
fish in terms of targeting foreign objects.
I am enclosing a summary of this program. Phase I of the
program described herein will correspond to "Phase II" of the program
described I however the tasks have been modified
somewhat to better serve the needs of the program. The remainder of
the present contract will be used to assess necessary parameters of the
electroreceptors system in Gymnarchus and Gnathonemus. Both of these
fishes would be used in the Phase I and Phase II of the program recom-
mended above.
I am also enclosing a more detailed description of the experi-
ments to be performed in Phase I of the proposed program and an
example of the type of experiments that are appropriate in Phase III.
Details on Phase U would depend strongly on the results of the Phase I
investigation. I hope that this proposal meets with your approval.
Please let me know if you can recommend any added experiments or
studies which would aid in the development of a prototype hardware
system.
With best wishes,
A PROPOSED THREE PHASE PROgRAM
TO DEFINE HARDWARE ANALOGUES
OF ELECTRIC FISH SENSOR SYSTEM
PHASE I
� ESTIMATING THE EFFECTIVENESS OF ELECTRIC
FISH IN DETECTING FOREIGN OBJECTS AT A DISTANCE
Perform psychophysiological experiments using different kinds
of electric fish in restricted water so as to estimate their ability to
identify the existence of foreign objects in this water taking into account
electrical discontinuities imposed by the boundaries of that water and
the foreign objects placed within it. Specifically, consider the fish as
a signal generator and receptor and the use of auxiliary signals which
replicate the fish's signals and other signals of specific interest. Com-
pute the expected behavior of such fish in open ocean or fresh waters
given typical boundary conditions with respect to depth, topology, in-
homogeneity of the water, temperature, and other applicable parameters.
Summarize these findings in terms of the ability of these fish to identify
foreign objects in a harbor or other natural body of water of interest...
foreign objects such as small submarines, torpedoes, scuba divers,
skin divers, mines, and others.
More explicitly, determine the sensory capability of individual
electric fish in terms of their ability to sense the existence of foreign
objects as a function of range, fundamental area, volumetric displace-
ment, differential discontinuity, and so forth. In this regard, use a tank
of water at measured temperature and electrolytic conditionts. Retain
a fish near one point and insert in the water various objects, discerning
the different behavior of the fish as these objects are inserted by con-
currently monitoring the electrical field within the water. Specific levels
of background noise will be introduced by using a noise generator and
measuring the noise amplitude in the tank. Experiments will be per-
formed under normal and extreme noise conditions. From these results,
1
calculate the estimated behavior of such a fish in detecting an object in
an infinite water domain and large scale waters with various boundary
conditions.
PHASE II
DETERMINING THE METHODS USED BY
ELECTRIC FISH FOR RANGING AND LOCATION
Perform detailed experiments wherein the particular character-
istics of electric fish are related to their abilities with respect to ranging
and location. Particular attention will be focused upon the use of phased
arrays of receptors, the fish's ability to determine incremental time lags
in the signal, the estimated spectral properties of the signal, the fish's
ability to modify the transmitted signal as a reflection of knowledge gained
from previous receptions, and so forth. Interpret these findings in terms
of specific schematics and data analysis required to synthesize models
of the fish's capability, models which when reified would provide signal
advantage over the state of the art with respect to such a target as described
above.
�
PHASE III
DESIGN OF EQUIPMENT SYSTEMS SUITABLE FOR
REPLICATING ELECTRIC FISH
IN TERMS OF TARGETING FOREIGN OBJECTS
Design, fabricate and test experimental apparatus suitable for
replicating the above-referenced models. Perform experiments with
this apparatus so as to improve its ability in various regards. Make
a specific comparison of this capability to that of an electric fish and
estimate the utility of such an apparatus in terms of operational situations.
2
COST ESTIMATE
PHASE I ' - $
PHASE It - $
PHASE III - $
6 months
1 year
1 year
3
I. WHAT IILVE LEARNED F1107...ri INVESTIGATION OF ELEcirni:
FISHES AND QUESTIONS WINCH STILL 17?3:1\ILIN
Electric fishes have specialized electric organs: (1) transmitting
7
organs which transmit electric signals modulated and coded in specific ways
and (2) receiving organs (electroreceptors) which are part of the lateralis
system and detect either (a) discontinuities in the electric field generated by
the trensnaitAng organs when the fishes -ose an active detecting system; 0))
changes in the electric field generated by the authorhythmic electroreeeptor
when the fishes use a passive detecting system and (e) electrical signals
emitted by fishes of the same species or of other species, even nonelectric
fish which still produce an electric field resulting from their muscular activ
during swimming. All this means that electric fishes can navigate, detect,
locate and identify living or nonorganic matter and can communicate under-
water by means of electric signals and/or electromagnetic detection. The
electric power involved is very small. Some of the electric fishes have
built-in a jamming avoidance system. Most of them can extract a signal fro;
a noise that is significantly higher than the signal itself.
- -
land from the data we have from other scient-
ists work, we concluded that unique methods of detection and communication
are peculiar to some electric fishes, which could significantly improve our
technology if properly applied.
In the appendices to this chapter I summarized the results of the
research I have done so far on electric fishes
OF C C OD.: --DLI
One hundred thirty-six fresh water and marine electric fishes were
. S2 fresh water fishes and 54 marine electric fishes. We
collected and correlated data scattered through magazines and the few
books having some cl-apters on electric fishes and data from our own
WE: iiStC:e. the locations of spec,-.:Ic anc: prepared
maps showing their habitats.
diseases of the electric fishes were studied. We found a
cure for the ulcer-like disease of E"ectrophorus (publ. Nature,
January 1964) (a cancer-type of skin disease).
'7'ne anatomy of the electric organs of Electrophorus, Malanterurus,
Steate,-,.e.nys, Szernarchus, and Cyrnnotus carapo was studied.
Histo2.0.7ical preparations were made oi' the C-leCtriC organs (I-Iernatox-y-lin-
Eosin staining) and of the electrorece-otors in the skin of Electrophorus
.schovsky-Gross silver impregnation with C.Ittjal counterstz.inia,;./.
Photorn:crographs were taken with a on -interference micro-
scope.
The type of encoding and modulation of the electric signals err'�ed by
zlectric fishes were studied and classific...d. ;Electrophorus electricus,
carapo, Sternarchus albifrons, Eigenmannia troscheli,
Cnathonc.imus peters ii, Gnathonemus curYirostris, Gymnarchus niloticus
electricus.)and a'o:eru:-us 'Information Processing by Electric 17'iShe.s'
:.-,ubliSh.e.d in the Proceedings of the 1934 Rochester Conference on Data
.::..ccuisition and Processing in Biology and Medicine, K. F.nslein editor,
Pe:-Eamon Press, 1965.
S. .7,-'ahs.'vior experiments were performed with Sternarchus albifrons,
Electrophor;.:s electricus, and Gymnarchus mloticus.
7. Proof was brought about the communication ability of electric fishes.
Their sensitivity to electric a.r.ci magnetic stimuli was checked.
S. Exact measl:rernents were made of their discharge; and curves were
traced for the Electrophorus establishing the voltage, peak power, and
average power of its discharge. Curves were traced also for voltage
against length of fishes and ag,ainst weight of fishes.
9. Experiments were performed, and the effect of dc currents on the electric
signals of electric fishes has been established for Electrophorus,
Sternarchus, Gymnarchus, Gnathor.emus, and Malar:terurus.
O. The effect of changes in temperature or he ta: water on the frequency
and amplitude of the signals e,--:"cd by the electric rihes was investi-
gated and curves traced for temperatures from 13�C to 28*(7..
"
The effect of anticholincesterase on the brain of Sternarchus wasinvcszi-
gat
The change in its behavior and frequency and amplitude of the
signal emitted by the fish were studied. .The acciiination effect at low
(23�C) and high (30�C) temperatures of the tank water on Sternarchus
albifronS was studied. Comparison with the anticholineesterase is made.
(Paper presented at the AAAS meeting, University of California in
Berkeley; coauthor with Dr. M. aslow, University of Hawaii. Only the
exeriments conducted in our laboratory are mentioned here. Dr. M.
Ezzlow conducted other experiments on Xillyfish and other fish at the
University of 1-lawaii, his results are presented as first part of the
paper presented 23 December .S5.)
A hypothesis about the way electroraceptors of the electric fishes work
was advanced, and experiments were c...z.-.vised to prove the hypothesis.
�Experizner.ts with moving eleetroltatic and magnetic fields were con-
ducted. Vie demonstrated that Gymnarchus niloticus is sensitive to a
potential gradient of about 0.03 ;,;.vicrr., and Sternarchus athifrons is
only half as sensitive.
oar
1=1. Alternative explanations of some previous experiments were given in
terms of this high dc sensitivity.
15� An exp12.nation in similar terms was given of experiments in which
Gymnarchus niloticus and Sternarchus albifrons were trained to
detect 0. stationary magnet.
13. The mechanisms available for the location of objects by electric fish ,
are reviewed; It is concluded from the results of a critical experiment
that C3ymnarchus niloticus can detect objects by the disturbance of its
own electric field in the water.
7. Lsr.tz.ns derivation of the approximate theory of this method of object
location is shown. The effect of the receptors of the perturbing field due
to an object depends on the electrical properties of the receotors. In �
cr:ztrerne cases, the stimulation of the receptors is proportional either
to the potential or to its second derivative. Graphs are given showing.
the effect of an object on the potential and on its second derivative around
the surface of the fish.
6
18. Experiments are described using Gymnarchus niloticus which (a) con-
firm that the mechanism of object location employs the detection of the
distortion of the electric field produced by discontinuities in this field,
and (b) indicate the limits of the sensitivity of the fish.
19. The detection of the second derivative mode of its own emitted signals
appears to be the most probable one operating in Gymnarchus. The ex-
perimentally determined limits of detection are discussed in relation to
the random noise in the receptors circuit. It is concluded that both spatial
and temporal integration are likely to be employed.
20. The. thresholds for object location and for response to direct currents
z...re. compared. It is cor.clucleci that the sante., receptors are probably
ooerating in both cases.
:Experiments were devised to find thc.1 attenuation and distortion of
signals similar to the ones emitted by electric fishes in underwater
is clearly p�I-oved that signals are distorted and their
shape is close to the derivative of the original signal. In case we can
devise �a receptor system that can integrate the second
Cerivative of the origina sal an improvement of signal-to-noise ratio
about10:: is to be expected. Experiments with e.lectrontagnetic fields,
pure magnetic fields, electric integrators, and magnetic receivers are
under.::ay. The results of these experiments may bring.us to the
reco,-----.enclation of an absolute new underwater transmitting-receiving
syLtern.
23. The electric organs of Sternarchus albifrons, a South American fresh
water weak electric fish, have been studied with emphasis on electro-
receptors. The morphological and physiological characteristics of
electroreceptors, ampullary and tuberous, were investigated. Special
instrumentation required for establishing the role of these electrorecep-
tors in object location, detection and identification has been developed.
24. We have recorded with microelectrodes the autonomous autorhythmic
electrical activity of the tonic asynchronous ampullary electroreceptors
of the South American weak fresh water electric fish, Sternarchus albifrons.
We have also recorded the electrical activity from the asynchronous
phasic tuberous electroreceptors and of the synchronous and asynchronous
ampullary electroreceptors of the same electric fish, Sternarchus
albifrons. Preliminary measurements have been made. The auto-
rhythmic activity of the ampullary electroreceptors has been demonstrated.
1
The electroreceptors are part of the complex lateralis line system of
the electric fishes.
25. The other lateralis line system sensory receptors, like mechanical
receptors and displacement receptors, have been discussed as part of
a general hybrid object detection, location and identification and recogni-
tion system of the fish.
26. A study of the anesthetizing effect of tricaine-methanesulfonate (MS222.
FINQUEL) on Sternarchus albifrons has been undertaken by plotting time
for the anesthesia and recovery for different specimens. The anaesthetic
does effect the pulse-rate and pulse-shape of the discharge of this fish.
Anaesthetic other than "Finquel" which does not affect the electric fish's
electric organ pulse repetition rate has been found. Also, the effect of
D-tubocurarine and the counter-effect of neostigmine has been assessed
for Sternarchus albifrons. Finally, some improvements in the micro-
electrode recording instrumentation have been made.
27. We obtained some specimens of the African weak fresh water electric
fish Gymnarchus niloticus. They are supposed to be the most sensitive
of all the weak electric fishes known. Together with two specimens about
one foot long, we received a number of baby Gymnarchus niloticus about
two inches long. The baby electric fish were infected with a Saprolegnia
fungus and could not be saved, but we fixed a number of them in buffered
formaldehyde and one of them has been cut and mounted in paraffin for
histological studies of the electric organs. Preliminary measurements
have been made on the communication capability of adult Gymnarchus
niloticus.
28. The electric discharge of Mala.pterurus electricus, an African fresh
water strong electric fish, has been measured in and out of water. Its
electric organ can discharge bursts of impulses of 100 to 350 volts and
currents to about 40 mA. In general they can put an electric power of
about 1000 watts per kilogram of electric tissue (1 watt electric power
per gram of electric tissue). From our investigation, it can be con-
cluded that electric fishes could use their electric organs (transmitting
and receiving) for navigation and communication � in other words, under-
water object detection, location and identification using an electromag-
netic system of detection.
30. The physical analogs of tonic and phasic electroreceptors have been
established. Both are represented by a generator connected to resis-
tances and capacitances in series and in parallel. The difference
8
between tonic and phasic electroreceptors is that the first ones have
one resistance in series with the generator whereas the phasic electro-
receptors have a capacitance. The tonic electroreceptors seem to be
predominant, maybe like five-to-one, compared to the phasic electro-
receptors. The electroreceptors seem to act, to a certain extent, in-
dependently of the main electric transmitting organ; at least two out of
three different types of electroreceptors are asynchronous and only one
type of electroreceptor will synchronize with the main electric organ.
It has been found that the complete denervation of the transmitting electric
organ does not stop the activity of the asynchronous electroreceptors (both
phasic and tonic). The fish is still capable of responding to conductive
and nonconductive objects placed near the fish's body. It may affect the
total capability in determining certain movements or impair, to a certain
extent, its sensitivity in object recognition. Some of the synchronous
tonic units are connected to one and the same nerve trunk part of the
acoustico.-lateralis system but connected to specialized big nuclei in the
brain.
The most striking fact about fresh water weak electric fish, besides
their spontaneous electric organ, is that all of them are provided with
a highly developed lateralis line system. Related to this acoustico-
lateralis system is an enlargement of the cerebellum, especially in
Gymnarchus niloticus and in mormyrids. The unusual importance of the
lateralis system in these fish, compared with other teleosts, is not due to an
increase number of "ordinary" lateral line sensory organs, but rather to
the existence of a great number of specialized sensory organs within this
same system. This is supporting our hypothesis about a hybrid complex
underwater object detection, location and identification system used by
electric fishes in recognition of prey, predators, and navigation in general.
It is recommended that the other lateralis line systems from different
fresh water weak electric fishes should be studied with the aim to find
out their role in object detection and navigation.
CONDITIONING TECHNIQUES APPLICABLE TO THE ELECTRIC
RESPONSE AND BEHAVIOR OF SELECTED ELECTRIC FISHES
Summary
A study described in this section is designed to identify the type of
conditioning techniques (operant or respondent) applicable to the electric
responses of selected electric fishes, with particular attention to those
procedures which can be used to assess the "dynamic range" of the re-
sponse with respect to major electrical parameters. Several additional
behavioral studies of electric fishes are also described in some detail.
Introduction
It was mentioned in previous sections of this report that certain fish
possess organs capable of generating electrical discharges; and, at least
some of these, have electric receptors that are capable of detecting and
discriminating among different patterns of discharge. This study is focused
on determining the e::tent to which the electrical responses of selected specie
can be brought under the control of respondent (classical, Pavlovian) and/or
operant (instrumental, Thoi:ndikian) conditioning techniques. (Skinner, 1
and Keller and Schoenfeld. )2 The results will then be examined to determine
whether or not significant relationships exist between the type of conditioning
operations which are applicable to a particular response and physiological
data concerning the structure of the organ, the type of tissue from which it
was probably derived, the type of neural innervation and control evidenced,
its electrical characteristics, and its functioning.
' These questions are of considerable interest in behavior science.
Viewed as operational definitions (Feig13 and Frank4) the two sets of con-
ditioning operations subsumed respectively by the labels, "operant condition-
ing" and "respondent conditioning", are clearly different. The controversy ii
the literature of the last decade or so concerning the kinds of learning which
exist has not usually been at the level of experimental procedures and re-
sults (i.e., of operational definitions) here assumed. Instead it has con-
cerned such questions as the possibility of reducing both types of conditioning
to a common, usually unobservable, intervening or mediating process; with
the nature of the postulated process; or, with the type of theoretical formula-
tion (Estes, Koch, et. al. )5 considered most promising by the particular
A.
10
-
Skinnerl� 6 for example, prefers the descriptive,
operaiional level and remains aloof fron-. ph,ological or phenomenologil. ca
as well as from the premature building of theoretical super-
struct-res. Hull, 7 on the other ':.and, had as his primary urpose in
writing his "...Principles of Behavior, " the integration of the two types of
' era:ions c.n 8 one k.
poszu_ate.s ri uncizrr...ying- process
consisting of ,13:-.1-7.;.1..us-r.::spo.--se associations resulting from
temporal contiguity, and suggests that both types of condizio---7-g follow front
- 10 � � v.. �-..r ����� ��� ;7. ���
� � i�ffe %.:4....%..���� \.....0.17:10..4.
". � � " "
ntc...dals as a i-r.c.lar.s of achicvincr the desired integration. Tolman, 11� 12
14 and, i:.a different. way, :NI:1'ler,
easizeoerce--)tual and cognitive processes as
Thera is no cuz.:stion, never, thc clear distinction
bz.-t7.ree.-. the two sets of cor.ditic-'-,g operations at the level of operational
in terms of laboratory :...-,rocedures and results. This is
experimentalis: does his work.
s:,ecies "-. . ...... Ia.:oratory, " two alrnoiE:
zitutually exclusive cl,....sses of responses can se oef:ned, corresponding to
40�46.0 .:0;01% U:4�:.*: set of condizio'-g operations which is effective.. That is to
ea :.:o..: are found to be conditionable by either responder.: cr .
'
-dre
�, usually not both. it has been further observed that
� ,
rises w��� � �����t
can be conditioned respondently typically involve autonomically
accessible to operant procedures
always ernploy stripe.d muscles under central nervous system control.
Tne empirical relationships between the two types of conditioning
pro diu-es are quite complex. Smooth =scle ("en-lotion:al") responses are
often cc:.c1-::::,ned re.spor.dently in the course.,of an operant conditioning program.
There alzo exists a small r.urnber of striped muscle responses which may be
elted respond:en:1y by noxious stimuli (e.g., shock-leg flexion) and whi�-...h
are reinforced cperan.tly by the terrninaticr..of, or escape from, the noxious
.,....Ithough the resoor..ses--in this case, le-,-flexion
,elicited by shock and leg-flexion as operantly reinforced--are superficially
the same, they are best regarded as two different responses with similar
����
topography.
11
������,�rew.,9.,
wii
ur.4er centr--.1 nervous syztem control and of respondent techniques
With C:Ontr011aCI, r.CS:D0r.SCS, is
. �
���,
,cs: L.,
.����,.� srr.coth muscle resnonses. Re:srdonder.t (Pavlovian)
tec't:es are an.olicable to all conditionable sr...00th .-nuscle responses and
� �
o e
which r.oxious e.licitir.g stimuli exist.
�� .0.00 at/ ..ro tor tto tn.. 4. tt� o. tr. ....ft. at � at � Iv 1.. too .� it.* � � t. � � � t. ot. . � otO
.�Z2
.., � ...1.- o�--���-� � - �-� ..-'
......... ........._. _.. ..erms ....,_ .16...s, ..,..,_ e...e.....1.;_e, most e....ectric organs derive. '
-...,- .
......,..... ....._ ....�......... .....,...,.....: ;...s.,........1 - �.._-___, ..............._-_, -..: ocular, deper.ding on th-
,-.... ---- _,......,�
spe5-- w.... ,...e ...,.....,.e exception of .....la...a.�-_,terurus, where origin is either
) � 17 albifrons is nervous tissue.
C .1Z � Z1 �
���� �� � ':ecn ...dr, .4.���������... � �.� ���� electric organ varies
spec:es, c..z5 G:
161 - t.: �
nerve,-_-..7, system of fish does not exhibit the sa:r.e degree of differentiation
����C......������C����= \ �
thes,...,: natural -..-ariat:or..L, seems worthwhile to
eznp-..r:eally the type of Conditioning oper:..--..ziz.�,ns vinich can be .used to t�;%����� � ���I�
�.0 \NO 41 ,a� .-.:s.pci.-.52s of several species, carefully selected on the bz.sis
of kr.own structure (and ava.ilability) and to re.late finding-s to the releva.r.;
information.
Sinee it would not be feasible to include in the investigation every
con-
ditioning ope.ra.tion in each of the two sets, emphasis should be placed on
those 7,rocz--..dures which can be used to assess the ndynz..,mic range" of the
. re.:ponse in each case. This will involve datermir.ing, not only the nne.an
and clistribution, but also the and minimum values of major response
-1-,ararneters, both in the free state and under attainable forms and degrees of
environmer.tal and behavioral control. aza obtained in this way can also be
correlz_ted'wit:- information al.-iout the strueture and functioning of the respective
electric: orgz.ns. Tl�-le data will also be rz..:levan: to the study of receptor
-processes.
is impt,rtant to study the types of conditioning techniques wlacn are
applicable to the electric responses of selected electric fishes, with some
emphasis on those procedures which can be used to assess the "dynamic
..� �
range o: the response in each case. Possible relationships between the type
of conditioning operations found applicable and the type of neuromuscular
�
12
, nz.tcd that ti.e relationshi..-;
discussed between the two clas,scs of conditioning a:1cl the two major subdivisions
of the neurornusc;u:ur systeni not an integral part of
behavior ..heory is::, or even Sklnner's version of it. it is rather at the
, sos.1 .16 ,���� � .0 D.. �...; at. 1.� ��;�,.� �;�� �.....C.: 4.1 � ��10 ����� � � ������������1,....
cf cnr.ditioning -.:rocedures (which are part of bcaavior theory) and a Sc: of
facts eoneerr.i.tg the c: the w:lich are CO:-.C.Y.7.:0i.k::Ci
facts are r.ot 7-art of behavior theory). Data pertaining to the "dynamic
range" will be c:xarnined in the light of known or readily ascertainable facts
elee-..ric. organs.
The behavioral research should' be e.:�:ecutea in two stages. The first
6. ..-.. which
��� �' ����". ��� �������, ����:'
,�������� dhd���� ��� �����, ���� ../4.6.���� �� �������� electric Se 1Z, S., � � INJ � ����� .� %1�1, fr." Na� ��� LI � � Sd
� II , . �
.3 "."-- S7,3:15 a.
cf Nind and Den-reeCA.73ehaviors.: Cc which is
�
��� � � � � � r � . 4.4. 6 C..ose..,.y related cues:ions
each s;ecies and each response:
� .
����� lb/. le � �; ���� NS SO 40 ���'r S.�;%.0:"
ar.d definable Class of ur.conditioned. stimuli?
�
the respor.sc-: a
?
r., the,. -4.e�-:-.3c...r.se be conditioned
le ���:%��.������� k. ��� �
�M,
to
re.
discoverz.ble
using res,-.)or.dent (classical, Pavlovi...n,
.�.8
oparar.t ( instrumental, �Thorndikian, .
prures employed to answer these cuestions will include the.
II... � �
� "� - C
a -
Ina spontaneous!, emissions of the fish have to be studied under
the range of environmental conditions temperature, light,ksound,
electromagnetic shielding, etc.) i which the subsequent conditioning
studies will be undertaken. (Ferster; 19 Schoenfeld, Antonitis,, and
20 e. r����
1 21) �a � distribution1
,��� The of all significant response �
parameters, including occurrence in te for the non-continuous case,
13
will be obtained. Since the responses of strong electric.. fish may
not occur as free operants (i. e., in the absence of specific
eliciting s'�:ll), the bulk of these data v.".:1 probably concern
parametric variations in the continuous emissions of weal: electric
come of the inform,tion :.7..ther,-..,.�1 at this stage will be uscful
..r. deterrnig the nature and degree of e.xperintental. control
owl �
necessary to ensure the
me.asurerne.nts.
2:rossor f.f.enc.:ral activity
.n quen;�.,"a� � 4.:1�7. C..
The "gross" or "general" activity of the fish has to be recorded on
a continuous basis in
�
which move.n-ient by the fish is possible, including the foregoir.g
ft � .
- � - o-_,eranz stucy. se: of floz.tz, and
a. as.
serve as the sensing device .Cr our-oose.
and st:--:,..lation has been asserted to ho7d for certain
3:
permit us to date=ine whether such a relationshi:p e:tists and,
so, to specify its precise form. Other uses of this measure will
be evident. as the. 37.Llay proceeds.
the effects on
to be presented-v.':11-..,e carefully selected I�epresc:rtt
....e range of modalities and parantetric values known or
the receptive ca.r..iabilitie.s of the organism or of speci.al
interest to the investigation. Variations in electrical response
ar,s.metz.,:rs, in the general activity measure,, and in other describable
behaviors will be observed and recorded. Other technicues (As'.123
- 24% � .
, C. the invest.:�raztor. . ot unconditionea
�
responc:ents (lc.urocia 25), associated with each s"�ulus and the effects
of stimulus presentations on a free operant other than the electrical
response.
One of the objectives of this step is the identification of neutral
stimuli which are suitable for use as "conditioned stimuli" in .,
respondent conditioning and as "discriminative s'��.uli." an "secondary
reinforcers" in operant conditioning. A sufficient number of `In:L.-nary
reinforcin5. stimuli" will be sought for use in manipulating the strength
of selected operants.
'14
proeedure--s provide the type and deg-ea
The we.-..k and 6.:eeies
in which frequency (anci pc..,rhapJ other parameters) is. said to
be highly varial.:le and upon uctivity and excitation
ly
o: the fish, aro. lihc.:.t.o conditionable operantly.
'7'he failure of 'Longo and to find inerez.sed resistance
. - - � " -
..� � ...�6
reinforcernent --a consistent finding in the a-'---1 laboratory--
illustrates the need for caution in generalizing re.sults over too
zeg:.nen. of the phylo_;ene-_-_.: sea_e and possibly over too
wide a ran-_-e experimentai operations, time schedules, etc.
�
%.� � .. � � �
- - 28 ����
gese %V V ��� � as "
sirnilari:3, of his results in conditio-:-g lever manipulation in
the 7.-1��".- 111 : �
sa - �v � �
�."-� sv :� " "�y �
v � . 4:: � ..1. .{. OZ.. �� ����
.
that operant conditioning rr.I.,:st be a very old and fundzIrnenza:,
tat..t....,J �
3aing, :: :s exteetea �����;��� � ;.:e ��� tl,"�������$ di,alb**V 1:0.
r.
species will far differer.ces. Nevertheless, it is
oartieular species,
..� U.... *. � 44.� \at.. S. � �:.� .as .� %a. 4,4.4. �
. �
,�����... � a� � to .4 I � ...a � a�
answers to

.
cb::ective of this stage Is to "push" resnonse nararneter.s, sihgly and
selected combinations., to their upper and lower limits.
is at this thz.: the diffe.rences between operant and respondent
con'iti.:ning bears ..7:1CS7. critically on :he oro�sosed research. Only operant
t f � f
O. 1.0 �
res-:,onses in the desired manner. The main procedure available for this
purpc.:se, namely "response differentiation, I.t consists in making reinforcement
upor. the en-.ission of responses progressively approximating the
desired values. (See for example, the oft-quoted study of :-.:ays and Woodbury,
initially reorted by Hull.) 7. - .10 'c, _1, _ '_ I. _ of _ _
cedues F er ster and Skinner 29) can be used to "compress" two or more
discrete resoonses
On the other hand, the control over the pz....ramcters of those responses
which can be conditioned only respondently, is generally limited to those
variations in magnitude and latency associated with changes in the intensity,
16
.3 I :'. r. toli f'
:,' :. t;
... ii c., 0 (.1 to
U) 0 :,' t . 'ii
fl et.; : ���
i; r'. ....., 1 1 r.
-',
- r. I (- s.� i 0 I " (.1i- . II '3
0 -
.,:� :,�' to Ii . : t.). :..).
( ) '
c i
!II (...� � �� 1`) :). i . tit
ir (!! c.. 6% i s (../ " 0
11 (... (.) . (l) ('' ;1 . I..
r.4 0 `, . : ' '
(.., � fel IN' G. --,
'S ���
el ,,
f� : , t-
e. (1.1 (i... fl ()
I) ; ()
8 c`' � ci ro :.3- 0 i j,(i
:3 0 f
.f......: :,' tr.
O , .. co $ s
'.''' 31 o c)
r: ...
:, to � I 0 IA si
el.
11 F) ('' 0
11 s... � 1, Ul 0 (1-1
et)
P 0 :3,. o r' ,
o. f.5. I
O 'ti ,r.;' c-.� il 1.1
:3 (4).�,f, En Co SI Co
Cl. C)� ' c 1'.
I-'. 0 r 1 W 'Li 1,1 1.. 0
C) , CO (1) r) t, .1
..:1 - , 0 . I
0 : �� Ul :1
et. �� �-.� 11
cc, C., 0 :r 1. .., . � � it P.' "
� ,
S.,/ (I: .f4 P CI le..
O e(,11
0. (,)
� 0 .-.3 *3 :, ,..�
4-. u,.
.., :-.
i 1 CI 0
+-A .0 < f 1 ...
� 4 1-: P'. IS :3 .13 In
i
0 0 f3'): � 0 0 i
tn k; r�. '6! 13 3.) t)
ri (U ;.!: 0
c, � -�
O I- .t 1 4. 1) rt . 0 1 ''
:3 r-l� r -�
� 1 ;-1 0
En 0 .0 0 0 0 1-� � �-4.
0 r. 0-
2, o ro ce ca p
Pt C)'
111
���� 1 4 r ,. , � Ul
-P. t , I' r,.
(1) 0 1.-.�
a� ' 0 ill
x 0 .1 �. ro o
$--.. (f; 1., � I- la 11 i�-��
'S
i� < o
O 3-- ) ::)
.., W . rt. ..., Ir I. fil V.
v... .. 0 - j
:1 0 ,,.. ,.-� (0 :11 0
0-) Ul �-�
0 . 0 r i :-4 C.'. '0
:3C'
iii� c) Cl' 0 0 8
' 0.. r:$ Psi :3 i i.
.. ,...
:i c,
.� ,... ,,...
.- 0 Lt.!. pl (3 .1 Cl
0. r.: �-� ,.; ;I :-.. :3
0 0 r) ()
tri .... � p�t). U: �, �'�
;.1 ;) ri- �t)
. 0.1 r: ..�
i� i�
n. ..,.
"3 frs
, .0$ 3: : (I. , ,.... 6 .( ;, ,., ,i rI ( ( i . : .) :: ' : . t . (, : . c : t P :
t ". :
0. 0
0. st.� 0 t : t :� ri .. 1 ' ('' (C.: f :. f.: :� 1:
.....
0 .� . 0 . ' - , ...,:: :. (.., i, .
...,.. f., ; r : ,
.: .
:,� � .., .; i; (.. c, i. r...::. c., (., f
.. � : 6,1 r � � '; 1. � � , � (` n. t$ :,'
( . � 1 .. I i .. ( . i .�
C
V? .' �7:1:.: I.: 1 .7� .... .14. ' '....(�..,' ..e.1- : �� (::
1 . . ... e ;
-LI ; i (..) to :� '' ; 0 ., V: P 1.1�:�, f�'.
: 1..: r) , � 1. ' S f �, 0
0
f
� : 0 :1 ./ i.: , (0 (,. �r r, ::. `C; :I.
I.;) IC . ih $:. �i, t'' 1) I'l el. :1 el. (;.!).
�� i � ,. � t.,
�L'e (1 ..,/ .� : �� r'l It$ SJI - : 1. I I : ' ��. i' I'll ���,. : .. , � � I i �." � -0'zs 0 (6' ri.�
r�� ( � 0 t( ,,' if, 0 0 I i' ' � $ , , 0 I: '/. (fli)
P) � s ,1 .. vs ,,, ,,, r i, .-,- (., t..,
t: '0 ..,,..: ::. :: :: .c: ..�: .:: t�C: :: .,�
,..
0 ...
0 1.. 0 . , .. . . r . r, _c, . . , ..
� It... �' ... : ! ',i,. () r, 6 P..
;:,;'..- r'.. r. (..... : ' i.;.� r'� i, I.
'v.:"... :';:::
. , 0 0 ; 1.. cf... 1.i !.'.;, ,.: ,' s: ucti, ,....
0 :) . (..., 34.: . : U. n c� � s o 0.:
I i O. 1 1, ta 1� � ; 1 . (...) , , 0
(..... 0 0 -...r. ., , if, :, .�,
. � ...., ..1 :, ., ,3 -,-, ..� . C.
(
0�.-- c,,. ' ��� ,. �.�., ... (�! 1 , a,
C) ,-�� ;.. . o i : I ,.; , .. �-., .� 4,
., o ri 0 , ,. :: .. ,, ,, t*,
(.11 ( I ( .1 . 1 . .. . .. , i .f.
Pi PS .(P
1 '� 1: :I ( j-,' () ... ' '� ./ : L, '..
(. �-.� ' i 1: (...: ol. ti ;:. *,�(� r�
-' r-t
, 3) ,, ) �� � VI � 0
1 o � e( I I :: 1 i , � � ' * �
I ; (It (C� )1 it. .3. ' ii: . II:: 1:::..... .1:(1,1 fill:i;:, Ikt.:. . P::: . ..!lit:
t�� 0 v, : i (..,:, r,, , . . �� . i . i� ' ', ,,
.0 c) , i�
0 el i 1
11 . � .... :.;. 'GI if. :(ki� !,... (... $(...., i!:),
ii. ; ,...,: ; ; c..... -., ,.. :-�: -1- '� :"
; i . � 0.
31) ,p. rof. it::. ;_.., ..1..., ,.....; ,,(i fr.)) f),:. '(tif, i. kil
� i r) c,,, $ � . z.�- 0
�I PS 1, 0 � ' ' 1- . 1') P. 1 ; (1)
. (1) f L '1 ' ,',- 11 r: 0 � ''''
i i .11 0 0 0 ' ' c). ,.. u* P 11.
0
. ,� t., ,i " �..
rl ;.�) ii;; :1 (..: :1,)
1 � W (`-.4 (0 % : .r.: , " i 11
..1 : In
,....1 CV r. (': " '-."'�C. n' �" ,c.
,. 1,, tfl ',i. p.. 1,1" ::.*:� `..:�� ::"-)' (c), In
(..; (1) C) P 1. ::" c-.4. till 0 � � .r1
u: ti :1' 3-* 0 : ' PI i:, C:
.i.1 .. tu r.) , �). .). :4, LI, , .., 0
.0 k< (.� P� (1111 c. :c � 1). "�.:*
cl 11) .0 :i i (;) tc- .' 0 .!. Os
(;:.3: (Ix' ; ''., ''(.1.:: ::01:: I:I.::: tr.,i :: ::'
, . ���� $ - :; 0 '14,*
i� � u) 0
(... .t.*: "
i . (., 0 �.. $ In
th
..� .
(..: ..,
r.
O *,
:1 0
� C.)
Pa el
O :)'
4 0
� � 0.
:t*
^ � :.1*
0 0
0 c'"
1:
I � --�
0. (33
(.4 � II
) ��� : . I .,
W 'i., P ' .I ::1 :. :
P r: .�
:."
�' : :* (3 !.:* fr �-.1 $) f�'
.. 0 's..� 0 ,3. V .-- 0
() ' .. 0 I
(,�!, rol. if, 3ts, c(1),. L... '-�1,
,.. .: is . ��,:, ..,
0 o to' f.:
.,� .i P. c., 0 �c)::: C.
r-' .... ' .� :' 0
Ii
rps; Ci), f:�.... 0, .0
� .��
�-� � 0 ,�� 0 ,.. 0
. . , .r� 0
roe. 6 ' " 3: 1- i -1.� . : i- I ;
- co vo-. P. f�
tj0 �
r! I�� C.- .. ,... co. cj
r) -� �
: f :
� .......... 0 C
< ;-� } < :),� C.a. :(i
f)
-)� r ::
.'
. (.� ,4 , .. .
����- �
�-���. �-
���� ....tar tat to C...a. itt.. a. ....V �At
rate--normallyl:ather � in fish-- elicited by a tail-snock
By airi.ig tne shock in proper ternp,-,ral
shi? with a fishir.g lure (SI...),a conc:itioned reflex _developed,
ate y at." tat:, � � � � � Y. � ��� ZW'
.��-� 1������ .
a-ALL&
Ow ;al/ is cece.eratIc,n in ne- art rate, Vor two.
nd ronze to shack. ;Zec.--:asa No:tern:an. 53)
of � ...
� � L: � .� � as � �
_
� a.
.... - S./"?
. %..t./..."" ��
. , a."
c .� � � 4.. %.� two .7.� .0 %./ � tro Va.,* as ....V �
With fish, there is the additional fact that the heart has only vagal i
in:_ervation.s one would expect, this response generaii::::3 to
second lure (L2), as shewn-or. the rig-.t side of the diagram. Ey
alone, i.e., without�shock, in a random sec.uence
with the regular conditioning trials, the conditioned response to 1,1
strer.a.her.ed while that to Lf; clizninisnes, and the fish comes to
respond discriminatively to the two lures.
�
The same basic procedure can be used to establish a respondent
discrimination between two F.;:irnuli differing only in some particular
parameter (difference thresholds) or bezwE2en stimulus-presence and
(absolute tsholds). ;For several examples,-
see Ash 23.)
Uncor.ditioned-responses to test stimuli have also been use... Xuroda25
loo.-ced for respiration changes in newts as an ir.dicatior. of-ser.sizivity
to such stimuli as pistol shots, t---;-g forks and whistles. Murphy
and Harris 42 investigated the pinna reflex to sound in the rat,. "with
at least as good repeatability as in.;the normal human audiogram."
24
fact that. rc:spor.dent tc:c�c.ucs can be used with constrained
� �.6.6...�������.14.4Ct.. P3.�
suitable for the study of electrical receptors. They enable con-
trcd1 over the of the fish with rc.s.spect to test objects
and eliminate chana--e; in electrical fields which :night be-aSsociat.:d
with mover:tent tltrou7h the tr-nk.
Z) �
As more is learned about the location and structure of the electrical
c:so: these :is:: and about the ,..f:erent paz.tways involvea,
��� � � I ��������1 %.,����������� ���� ..11�0./ .0%) 41,06�� � ���
to verify the function of the suspected receptor and to cstirnate
LtUdi.:S '� � � � � �� - � 54
o: .narid in 55
lebes in the chick
, and thz: of :0' 2.
4.... 4.4: 4... 6.... ..
Locs.tion of Electrical Rece7)tors
���� �
.; aer �
the University of 17-lochestar,
Yd. ...Y... 6.0 ;40 6. 1� � � ���� � ka; A. km. 10 � � ha, ea 4���
pro-
fi.:h is dised
.
�
����
' � ' � - ��7
^ numerous co:lc-it:0 �.d."C Co
ow � � 6.6.1 0interesting:
�
entsnecessz...ry to clarify the results of the main investiga-
tion, including those control studies tner.tioned in the following -
section on 17-.'7,:-.)erirnantal De.signs.
irnents which repeat, with the species arid responses included
� main investigation, a good�cross-section of the prototypical
con-%,:cte-1 in the 11.,n.irna: laboratory." The results'
would provide additional data on the phylogenetic generality of the
conditioning operations, data and relationships obtained with other,
usually higher, organisms. �
25 .
C.N reliz...bility of data. related to all significant conclusions has to
ec estz...blished appropriately, using the bc-;st tec'�'cLues available in each set
� . � � ���:. �����;-� ...AA: S., of pruce.s
identcation of the relevant variables, a proper choice
of ;he::easure:is-, to be -used to represent -the dependent and independent
selection of variables to be controlled or randornized, the removal
effects, and other experimer.tal procedures and techniques,
desi-_-ned to mize error variar.c,s and to rnove bias or con-.
z:ne-.-,roblents can be anticipated along these lines. For example, with
spee.es, siee and prcicurentent will probably restrict
It l�
setc-aets ' which can be used in a given experiment. This will
S.: r . a.. �ao ��.: . a. S., 4.. A. ..... V ......, 4,06 A � . . . a,. u rk.) a. .�...a../ V. ..... A. . V .1. , S. s , ,B. . S S. . a.... L: s .. S , s , � � ��� ..: .. . � J �,... -
� � � �
.................
�
...1*.XV��
........:. -.....,..:. � '........
..
4".., .;AUS.1:-.).;::/,
�.C�
..��
�:��
,r.) ������ � .............%-.� �

,..1'4',. ,s. ,a. Z.,
t... - ,Ss,:... ,�., So:. i......., .......,,,�... ,.... e........ 1/4... ........
%��:$ '.., wo. s , , � A a ...,_erni-.:,-;..-..:-,-.--ne, 61 .,.....
.....,,..:
��
.� . 62 -
necesear;
sua:-. eireurnstances to shift emphasis to
each individual to serve as his own control, to
koms ��.�:;�Ca tte.6..��������:�������14,anc Ukd 0.0�;� 110415
are found. in. Skinn.e.-r, 1 Ferster and Skinner,29 -
ar.d
63 %T.-..:never feasible, strong preferer.ce will, cf course, e accorded
'hose which assir:n probabilistic warrantees to _ e.e
.�,or &MP � SO1 ���� NM. OD � � 1 S. �
problem which may be encountered failure to achie.ve any form
of control over the occurrence of, or over the parametric variations in, the
electriczil response of a given species. This would represent the most
outcome to interpret and would raise such questions as: Were the reinforce-
meats, other stimuli, etc., appropriate to the organism? Was the emotional
or physical condition of the organism such as to adversely affect lez.rr'�-g?
Was, the tk set to the organism too difficult would some other choice
have -rforl-ted out differently? Were the enviror.Mental conditions and other
laboratory arrangements unsuitable in any way? Were the conditioning- technizlues
competently Etc.
ObServazion of -..ha behavior in the course of
conditionir.gs. will usually provide some relevant evidence.
:a addition, it will be highly desirable. with every species, to select
convenient respondent (e.g. , McCleary64) . and a convenient operant response
Bently, 65 Deterline 66) to serve as a kind of control. Then, if a
particular conditior'^g paradigm fails to yield the expected results, as applied
26
�
to the c:loctrical respons, the ::rocenc,ures usez: e.'zJ;.�:k1 out
(I
in st'udies of the
electrical receptor, z.n alternate set of non-c1c.ctrica1 discriminative stimuli
7.1 S a check on procedures Nt,hca needec.:. )
-nror t: co o.ncr means, the kind of renroducibilitv and pro-
� ..� �
cision in results I:I characterizes the best scientific resez.-rch in the
A.A.A.
2.7
References
1. B. F. Skinner, "The Behavior of Organisms, Appleton-Century-Crofts,
New York, pp. 457 (1938).
2. F. S. Keller, and W. N. Schoenfeld, "Principles of Psychology, Appleton-
Century-Crofts, Inc., pp. 431 (1950).
3. H. FeiE,r1, "Operationism and Scientific Method, Psychol. Review, 5,
(Symposium on Operationism), 250-259 (1945).
4. P. Frank, Foundations of Physics, International Encyclopedia of Unified
Science, Vol. 1, No. 7, 1-78 (1946).
5. W. K. Estes, S. Koch, K. MacCorquodale, et al., Modern Learning
Theory, Appleton-Century-Crofts, Inc., New York, pp. 379 (1954).
6.
7.
8.
� � .� . � �� � tr. v..... � � to vo.
1.e.; so, 70 �
�
���� � � .4. 4.�
-2 � .7. e ��
� � � � � 4..
"
�:1 V �
7 _
���������� 2��� �
11 -
- _ � ��� .� � ���
r � � a.� � � ..-1.��� � :a, �Lor:�� ��� r�I, � � ..."����; � ��� � a."... � � .� � �� � ra � :1 1.. � \�� � c � t� ���
�: - _
so � � ��� ���4 � a� 6/... a lob.. ���
.2:"1"e SS, �=4C�l�
�
...��������� row.... � � ������ r 7
9. Guthrie, "The. .7.--sychology of Learning. New Yoric:
pp. 32.0 1r.;52 (also 10351.
10.
11.
_ 11 rr,
. � � ���I C. � tor � :L.' 1..U.If /7:6' an. ��� le. � 4. r;LA. *yd. Co; a .�
Ty of
� '���� � l� � 4.4 a. P lo �a:* so ����� �-� � .....
"Y .., pp. 453,
1: -
adi.40.��2. �A, MID C.' i���
12. _7. C. Ton, i'v":�-os in :::;..c.ts and Men,"
. 7 :" � � �� r �
..Jva
13. R. S. Wooclworta, "Reir,foccr.az...nt of a Perc....cption," Am. 3. Psvc.c1cgy,
1.4
" r, o.; iv, �
%,....2%;� I.
A
�
14.
G. 1/4.:1 ^ - -:1 C" � and M. E. -7itterrna..-., "Reinforcement and Learning: The
Li
':),::�SS of Se:nsz.)---y 56',.25.)2:303:(-19).
28.
15. r. � ;��� ,
�:��� ��� � � � so � ���.: � *CI � k�Z 4...����� ioe .� nc: � .Prihran, "Plans ar.d the Structure
of 22havior," Henry Holt & Co., New York, pp. 226, (1960).
16. C.Zr-ndfest, "Irilectric Organ", In McGraw-Hill Encyclopedia of Science
and Technology, Vol. 4., 427-132, (Authorship inferred frorn the initials
....� ... � /
�
�
17. D. :eyes, "Electric Organs, " In Margaret E. BroNvn,-"The Pysiolq
of "� Vol. 2, Academic Press, Inc., New York, pp. 323-343, (1937).
13.
C. 1 11����,L E. '0 Tr,
�-� � �-�:: � .1...e Nervous ....a � � � ea a a �.:�. a ���I � 4..1 a� V � ���
ic,2.0:�, of Fishes,ft Vol. 2, Academic Press, New York, pp. 1-119,
19.
���� ^ � � �� ; a�-% 1
�� � � a S./ Sa t� Z."
20.
2
; � � �A ��� ��� 1 ;Z..' � 11 ^IA
50,
- �
.,.:�,�erant in the. Analysis of :BehaviorII
'see
�!` � " � � .1 "" P: � � ,�� ���7 """) � i
eno � � � � V 'v., � �.� .4 � ���� ...a � W d� � � .1 � %ors ����� ft.. � V � .��� ���� 1,1 ��ly U. �%.� an. �Ani �.�
� ,�� Zs) / Lo0 ./...��;741 ,1�� 4���� � l/ 1.� � i�� � le � � �� 4��� d� � ��� ���� Co� 1.. �.� �:11.1 gno. till% � .46 1.4�D ly �
. . � , I �...
�:
V la .0 ��� ".��: � i�
I en.
'����' � amm, ��� %�.� � ��� .� � aCi � � ��� a � 4. 4o...4V ���lo ��� �p. �� � ������� ���� ��;7� �I � ii�'� � 4: 31:4::�.� 4110 41740a
_
� i� ��� SI'�..) �
A c:: Press, Inc., NeW Y 0
23. 'Is% - ' �7� 1.7:1=f-3n.
......�,.��������4 1. a a � �� ���� � ��� ��
��
24.
25_
�
� e", � �Im ��� SVC. Id 'C. -0�� . leo..
. 1 � ,
4 r., 4S,
2 E: � �3 2 S 9 5 1
��
�
i� � ���� � �� aa��/��
.� � ��� � s�
kr� 6,�� � ...����� C-...
�
s���
1�� ��
�
oeelee
enee�������4�./. al� � ;,:A�
%ZIA 1
�
� ,L.,
��� � .4�01.)..,
T�":
���� � ��� ��
�� to so. ere ���
6Z, I.
����� 'we �ZO � �
to � � ,Wr
��� � wt.
sr/ 4:����11 ���
01.
km, ���
kr .4:01���
���,.; � �
��
Amid
� � g�
.41 LI' A,
� �����. _���� ��������...
me � � ...aloe.. sof
.0..�����
11/2'
��
7
V lo �;���
�-�*.
���� �LON
j �.����
����� ho�
.11. � %Of .;;�?
����
����� ����
�� � 0. L.) 0.16�
�
��� rl
1...� ����
�����
41,11,1,��
I
r r��������.
�����
�������
T
�
�76....� �
�
V:
N.
����,.
�,.00
's
mop .��� 1�./ /
�
�
26, V. L. . Bennett, ar.d Grundf,:::St , .1. Car� Phyz.4.ici.,
q* f
it) i k����iiji�
27.
"vr -, r-
...�� �-d �
C.:T. 0: r
P.%.:.c-Lic.a onaesiz,zance to 1'x:inc.:ion in the
169-172, (1960).
7 t'--...---..-
28. P.2. Dews, Il
Some Observations on an Operant in the Octopus," J. of the
0.7 7:.:.-hr.vior,2, 37-33.
29. C. L. .7erste-r, and I% F. Sinner, "Schedules of il.c....ir.forcernent,"AtrAeton-
...-sentury-Crofts, Inc., pp. 741, (19.57).
30. . 1-.7..luard, and D. C. Marquis, "Conditioning and Learning",
"L. C,3=pnny,
�-�� �
31. a in theYLe University Press, New
.�r: rk=4.7.1
�
32. :.�- 1 9
23,
� :32-9Z3,
Nor. Id S./ I.
33. L'...ssman, and "int: Mechanism of.0bect Locaz.ior.
I � ��� � V. � � � � 4. tOt .1�� 1,����1 401.1: _ s 5 Fish" ..T�
(195L).
34.
Nw�� n 'NO c.��� -
� 'ID � � � IS � �%.� 1 ; � ���� I. 4.1
�
m...������� " 77,
35. _..� Pitch
and Loudness 72hre.shoicis in the White Rat" J.
IT
36. :� Press,
v, (5.V. translation) (10,S7).
I � � . � �
�� 7 � 7
37. .-. :ha 1.7.se of S-oeci:ic
.3 z.epel. From Cdbjectiona'ole F.ocsts,"
4.V 4:1 a. :1...)::1�
33. ac o.-.s to the Pain of Others",
..'
� - _ - -0 ,
.fi. 464 41.1 �a 1 �
30
39.
�
, � � � 0.
i ����
ie
fLd,
11 , _
io.Ex-oar:',:z.11 Arysrc.,-ch to Arxiety, Escape and
i.voi,Lance "An:iety, " Ed. by?. a loch J. Zubin,
. -�
��� �� ����, � sr .r�
41. V.'.L.Deterlino., ItCperant Dizeriraination Reversals: Comp.s..rative Data,"
.7. :::::-z..-;rirnental ..knalvsisof73::havior, 3, :1-17-253, (1930).
42.
43.
J.., � ^-d J. D. :
7 ;
2,1%... � 1 da � 4,2 1 J. zie z. t,
- �
�
� � I.
"Neg1ib.a Effects of (.6X-Rdtio�
The
J. of Auditory Research, 1,
I I I
. � ye � arolCor......01 46. , � Gr. C 6..� ;
Lura:ion- Szu.;:y of Secondary Motivation and Rawarc.',
� 54,, No. 5, pp.
CA
....c
kJ.
44. IN:. "A Te:-..t of a Theory .of Shape Discrirair.azion
. " C.:c;-.-.ive and ..?...-.-fsiolozical ?svc.r.olo..7v, 52,
45. c.:1;. Sutherland,
46.
47.
4
1
o'
� ---������ ��� �� 1100.,�.,���������� %���� gr
�
... �
C���� ��� It .7 � ..��� �61 � '7". it � � �
1.0 .,��� � � � Ld ��
and Co., New York, pp, 943, (:954). A substantiai revision of
..k.,o.::�aorth 1933 ed. of Experf.rnental Psycholo;y.
�
r
:�� �
�
� .�
� .o. a. , Zap, Co. ��� � � � � ....Co... � ��g; �
16.0 �
�
� � � A., AA.. � ��� � 1 .0; � %ow'," ��� � a. � � ft mob...6*A. O..
CO a
" . � .
40. ..., .6.� ���� �� ��
� � . � ������
...alga � .���� ���� ��, I ZJXnerinnC:�7:�:11 A.ivs.so. .:..-enavicr, 1, 31-43,
50. C. C.101:-..-avit.c..h, .
7.4 � ..
A . � a... a
T...
�.� &�
11Auditory Thresholds
31
51.
52.
53.
��� � �
...: �
�.������ fd � a. �
� -, ���� _ 11 ,
� � 4. a. ve., Xi
.23, :.:13-123, (2.943).
1 t �_ � �_� � � � � �
�-� � �� .
���������.�
of Con-_-,--ntive-
.�
�
0.? � �
4.0." � ����� � ���� ��70 Loa. ����.1.�����:�%%/��,
-yr c - 7 " . D � Bersh,
.
��� �
Z. r)
"Con'.itioned
Rate Response in Human Beings D'uring Anxiety,
V so �����,� ��?vsu I, � ,
7:
54. 2. C. 1,7/ever, an.d J. A. Vernon, "The Problem of 1-:earir.;-;
C. o.. vResearch. 1, 77-5.3, Z1961).
: 11���
.4. 1.0 ��������%��
55. 2. C. Weyer, and J. A. Vernon, aring in the Bat, Myotis
ti
�������.� �� �� Cochlear The td J. of
.� � 41, 1.�����
"r �
,
��=3
56. J. Peters, A. R. Vonderahe, and T.:. I-.I. Powers, .Q.Lectrical Studies
c,2 1���\:: ����������.�
� enta..1 Zoolo72-v, ("LS;53).
57. .��� � ���� 1,��� 11... � �
4 . � � �� � � � � � .;�� � ,� Gt. � 1/4.*: � � CO � � nta .1 21 a SI ..;r�S o.; Co L... Wsy
� - -
��� ��� �4 ... . �J.., � so v 1 �
1 � � ... 1 � -
53.o ... � ..
New York, pp.;4i3,�.1.:".�.+ � j�
59. A. I. dw,rds, "Experimantal Design in Psychological Research,"
:-:o1:, Rinehart Wiston, Inc. New York , p. 446, 1950 (also 1960).
60. C. Clc.,�uld,:�.n, "Methods of :--t.,,tistical Analysis," John Wiley & Sons,
evF York, pp. 457, 1952 (s.-iso 1939).
61_
� �
New
� -
46 4. .���
A1.4 .1.1� OL. � � �.46 � � ZO �::*
-.- '7' � en �
�
I
John ;i'lley
62.� - �
. �
�����:� O. � .17, 4.71. ������� �����.. .11�" M. ZIO Ap OM, ��� � %:.� � � ����
r e�t
���
63.
1������ �
� ....
;-.:on
0:� " �
I 1/4 ��� I �
- f� � - :
. ...
Cc., Bos::-.)r., pp. 393,
���� � II
.a. � k,�� %.� \ � �,
Basic
4.
�
Nc.,w
32
� two. flow '75" Interocu.lar
"
� ...�
� '.'d).
II a. �
�������.
65� �4
:
66. L. ^ vafc.:.- o,
(1C:J56).
^.7."
---- :77. � L.' �
PROPOSED BEHAVIORAL STUDY OF ELECTRIC FISHES WITH
EMPHASIS ON OBJECT DETECTION, LOCATION, AND IDENTIFI-
CATION (PHASE I)
Three different kind of experiments are proposed with the aim
to assess the ability of electric fishes to detect, locate and identify objects
underwater. The first system is designed to use as a subject, a weak
fresh-water African electric fish Gymnarchus niloticus. A maze has been
designed (Figure 1) to be built in a water tank of 12 feet in diameter and
4 feet in height. The maze has two channels supported by bars on the tank.
The fish located at C will be presented with a stimulus in the form of
objects (metallic, nonmetallic, magnetic and nonmagnetic) of different
sizes either moved from D to E or from E to D. The reaction of the fish
and its choice of the channel will indicate how well the subject would detect
different objects of different textures and sizes. The fishes are freely
swimming. The second system, (Figure 2) is designed to prove the ability
of electric fishes to avoid obstacles like fine aluminum wire or nylon thread.
The African electric fish Gymnarchus niloticus and the South-American
electric fish Sternarchus albifrons would be used. Both fishes are blind and
use their electric transmitting,-receiving system for navigation. Gymnqrchus
has a very steady frequency (around 300 Hz) and Sternarchus has a steady
frequency (around 750 Hz) provided the water temperature is held constant.
The fishes would be presented with stimuli (attractants and repellents) and
their behavior will be monitored by filming it. The fishes are freely swim-
ming. The third system is designed for variable rate of pulses electric
fish species like the South American fish Gymnotus carapo or the African
fish Gnathonemus petersii. The subject will be confined to plexiglass tube
with holes and two electrodes (at the head and at the tail). The impulses
emitted by these fishes will be amplified and monitored witI2 an oscillo-
scope and a frequency counter. Objects (metallic and nonmetallic) of
34
different sizes would be presented from different distances and their effect
on the fishes pulse rate will be recorded. Graphs will be plotted relating
material, size, and di.stance to the pulse rate as compared with the steady-
state pulse rate. These fishes increase the pulse rate when an object
disturbs their electric field generated by the transmitting organ and re-
ceived by the electroreceptors in the skin of the fish.
It is hoped that these three sets of experiments will show the
sensitivity and range of detection of different objects by electric fishes.
4.
35
1 �
$ome .7�4
�
. .. ..�_.
1 I .
; .
1
t
,S i
i I
1
1
.. 1_ . ..0 � ....
,
02,Ft .:ssoot
001. '
Lvh rg.
tMZE
Maze for Behavior and Object Location Experiments with Electric Fishes
Agrik FCA,C1011{LRC
IsPt WI V
-
IP, Ft dtahl..
3!.;e10 rest. ki4tCp./.414
for bth vice eepfrinirit.is
p.
�
Wire Maze for Behavioral Experiments and Object Avoidance by Electric Fishes
�
dyLo kly4
� 1!1* ft 3 00 Cot. immx
Th !pc
13E (Rd tit.13)
EIE-CT rt 1 moil a
).
Aize.$ FoR'.
r ,
s c oPe
. . . . .
� � Experimental Set-Up for Variable Frequency Electric Fishes (Gymnotus ca.rapo, Gnathonemus
IV. AN EXAMPLE OF HARDWARE ANALOGUE EXPERIMENT TO
SIMULATE OBJECT DETECTION BY ELECTRIC FISH (PHASE III)
It is possible to simulate an equivalent sensory system of electric
fishes responding to different stimuli underwater. A system with a double
feedback mechanism can be envisaged: (1) one represented by a constant
frequency electric field transmitting system operating on the phase-syn-
chronous electroreceptors responding to discontinuities in the electric
field or to changes in the phase relationship transmitter-receptor; and
(2) another one represented by a variable frequency transmitting system
responding to disturbances in the field between transmitter and receptor
with a change of the frequency of the transmitting electric organ. To this
we could add an independent dual autorhythmic receptor system: (a) re-
sponding with the increase or decrease of the authorhythmic frequency
depending on movement direction of the disturbance in the electric field;
and (b) responding with a change in the latency depending on the magnitude
of the disturbance, and also distinguishing between conductive and non-
conductive objects.
A simulation of the electric fish object detection system would
make it possible to find out how models of the physical analogs of the
sensors could be integrated in object or organism location, detection
and identification. The range and sensitivity of the system could be assess-
ed and improvements could be made. As a first step toward this simulation
we present a measurement plan for an experiment in Phase III of our
investigation, to build and check the first of the above mentioned mech-
anisms for detecting, locating, and identifying objects underwater by
means of a phase detector system sensitive to changes produced by dis-
continuities in the electromagnetic field.
Systems and equipment for detection and location of objects in a sea-
water or freshwater medium have to be designed from the standpoint that the
39
medium is lossy and has a high dielectric constant. Consequently, the
wavelength of a signal transmitted through the water medium is different
that that of a signal of. the same frequency transmitted through free space.
Attenuation and scattering are also different from the values for free space.
The advantage of using a phase comparison, position determin-
ing system is evident because the precision of our measurements depend
on the phase measurement precision capabilities independently of the
frequency used. This gives us the possibility of using low frequency com-
bined with a highly accurate phase measurement.
In free space we have the well-known relation between the speed
of light c, the frequency of an electromagnetic wave being propagated f,
and its wavelength
c = f � A (for all A in free space)
In seawater, this formula cannot be used because the propagation
velocity of an electromagnetic wave is different than the value of c in free
space. Moreover, the propagation velocity in water changes with frequency
and is not a constant. It is represented by the relation
s = F 'X' where S = Function (A2)
** s = speed of propagation in water dependent on the electrolyte medium.
I = frequency (Hz)
A' = wavelength in water (meters)
The signal travels a distance D in the time T where:
T= D
T
* Taken from D. L. Nichols reports from the U.S. Navy Underwater
Sound Laboratory, New London, Cormecticut.
40
sca,,,vre,on- �
s.is dependent of and A!. With respect to time zero the phase of the
t-a.ted signal advances through the angle 217 IT. The received signal tlre'refore
; behand the signal at the signal generator by 277 fT. If this phase lag could
measured, the range could be calculated. The basic equation for phase-
rIparison distance measurement is
p =27r iT = 27rf
,ere p is the measured phase difference between the reference signal and the
gnal which has traversed the distance to be measured. In terms of the
.stance traveled,
D =
SP PA'
277f 277
)hase measurement can be based on: (1) the multiplication of two cosines
or sines) and integration over one period of the function; or, (2) by determining
the instants at which the reference and text signal cross zero in the same direction
and measuring the time be tween the zero crossings. The phase angle is then
2771 T, where T is the time between zero crossings. .
We have chosen some hypothetical parameters for a phase disturbance
detection and location system. The values chosen may have to be modified
according to our findings during checking, experimenting, and making measure-
ments in a scaled model in fresh water, sea water, and mixed fresh and seawater.
The parameters for our system that have to be scaled down are:
Transmitter: f = frequency = chosen between 5000 Hz and 1000 Hz
= length of the antenna made of a water column in sea water =
between 22 m and 50 m
p = endplates for the antenna feeding between 25 and 50 cm diam.
silver silver chloride plates on Monel-metal 1/S in. thickness;
plates will be fed by coaxial cable in a coppertube 1/2 in. diam.,
1/16 in. wall thickness sprayed with a teflon coating of minimum 1/32 in thickness.
Transmitter power effective delivered at the endplates = 1 kW. Transmitter
41
be in a screened room, double copper mesh, with a good ground to an
endent ground. Transmitter has to be crystal controlled.
.ntenna depth = 3 to 5 m under the surface of sea level.
teceivers: Instead of clipok antennas, ]ow-loss toroids will be used, tuned
to between 5000 Hz and 1000 Hz frequency fed into matched lines
and to low-noise linear amplifiers. The amplifiers have to be non-
phase distort type. Phase active networks connected to the ampli-
fier will correct for exact 1800 phase opposition of the toroids'
output. The output from the phase-corrective networks will be
fed into a variable high-gain (decade amplifier) differential
amplifier where the output will be monitored by a scope and a
null detector. Differences of 0.001% may be possible to be
detected at the highest gain.
Scaled Model:
George Swain* tried-an experimental toroidal antenna
scaled to small diameters for frequencies of 1.5,
2.0, 3.0, 4.0 and 7. 0 MHz in a solution of NaC1
of 0.285 moles per liter and a conductivity of
2.8 mhos per meter at 250 C in a tank 3.5 m in
diameter and .55 cm deep.
George R. Swain: Antennas in or at the Surface of a Conducting Medium at
'LI-, Report EE-116 of the E. E. S. Univ. of New Mexico, Albaquerque,
M. (1964).
4.
42
� .
t is proposed to use a plastic tank 12 ft. in diameter and 4 ft. deep fo7
ed down experiment with fresh and seawater.
We will increase the frequency to 400 k z corresponding to a wavelength
vater of 2.25 m; also ten times smaller than one of the hypothetical values.
z. endplates of the antenna will be 8 cm in diameter (about 1/10 of the surface
the hypothetical plates). The power of the transmitter will be decreased
100 W at the end plates. Antenna should be submerged to a depth of from
to 50 cm.
Receiver antenna will be made of corresponding toroids for the frequency
E 400 kHz. The ratio of the attenuation between a signal of 5000 Hz at 22 m
nd 400 KHz at 2.2 rn distance will be calculated and the factor applied to the
�esults of the measurement.
A piece of cotton filled, cotton covered, and partly isolated cotton body
with a volume of one-tenth that of a human body and with similar conductance
will be used to find out the phase deviation produced in the field in different
positions between the transmitter and the receiver antennas at different depths.
A device for making artificial measurable waves on the water surface
will be used to produce scaled down waves.
A random-noise generator will be used.to simulate natural noise which
will be added to the signal for different S/N ratios. Polarization and position
change of the toroids will be checked to find out if it is possible to reduce
the unwanted effects of water movement and noise.
The temperature and the salinity with respect to the conductivity of the
water will be continuously monitored and recorded during experiments.
Receiver and transmitter antenna will be checked under steady positions
and small movements to determine the effect of moving the antennas on the
�
43
�
3nse of the system. Q-factor of the toroids in air, fresh water, and
-ater will, be measured. The ratio between deviation produced by waves
or noise and the one produced by the object will be determined in order
ecide if a full-scale experiment is warranted.
fLATE I/S,
MILL
ar7:z7)1.;;
PLATE Ao
Simplified Block-Diagram of an Underwater Phase-Detector
System Used to Detect and Locate Objects
44
.611
 tIOM
Ratio
Divider
Example of a simple ratio .bridge.
45
��������
Impedance
ronitoring
Uct.:;ork
Trnnomitter
100W
01�1.1.1m.14 ��������� .1104������� 01=�������
������������
Oocilloocope
Phase
Correct in
Network
2.2m
d
el M O..%
4.4
Ag Ag Cl
Monol Metal Electrodes
2.2m
Water Tnnh
3-5m in diameter
GO to 150m deep
made of plastic
Toroids
g:- 1
Phase Correcting Network
D-A
g 0.1, 1, to; 100; 1000
Null Detector
Sealed Down 1!..1.01 for Experimenting Purpose of n Phase Discrimination
Object L.:tection and Location System in Fresh and Seawater
A1
Cathode
Follower Filter
r>
�-
`;tt.:3J
Operational.
Amplifier (Gain 100)
Operational
Amplifier (Gain 100)
. 180
Range
Multiplier
B+.
, hP
A3
I Calibrated
Amplifier ; Phase Shift
VIMMIII.1�11�411�111MW
NE 51
Overload
Alarm
..m.��������������,
Differential
Amplifier
Null
Bridge
Range . Cathode Filter Range Amplifier Calibrated
Multiplier Follower F2 Multiplier A4 Phase Shift
M2 A2 M4 Ph2
Proposed experimental set-up for demonstration of the "System for Detection
and Location of Objects in a Seawater or Freshwater Medium."