Measurement of the body surfacephysiologic tremor or "mlcrovlbratlon"

ALLAN F. PACELA,1 BECKMAN INSTRUMENTS, INC.,APPliED MEDICAL RESEARCH LABORATORY, 2500HarborBlvd., FuUerton, Colif. 92634

Five :sensor techniques wereevaluated for the measurement ofhuman body surface vibrations: a nonconmcting ctzpaeitoncesensor, :seismic (velocity) and rellltive electrodynamic types,linetll'-variable-differentiJll transformer :sensors, and the minllltureaccelerometer type. Of these, only the miniature accelerometerproved to be entirely satisfactory. Theelectrodynamic transducerutilized by several previous workers was found to beunSiltisfactory for use asa physiological vibration sensor, due toresonance problems.

Although this study was primarily concerned with instrumenta­tion techniques, preliminary observations indicated that theresting hwnan microvibration contains a large cardiovascularcomponent, probably obscured in previous observationswith theelectrodynamic velocity sensor.

Human microvibrations (MY) are normal or physiologic bodysurface vibrations of micron amplitude, observed at rest. Themicrovibration has also been termed the normal or minor tremor.Measurement of such vibrations is of significance in study ofParkinsonism and the Intention Tremor (Wachs& Bushes, 1961),and has been used in experimental psychology to distinguishneurotic, schizophrenic and control Ss (Williams, 1964a).Microvibrations may be described in terms of displacement, or itsderivatives as: microdisplacement, microvelocity, and microac­celeration. Although early research emphasized this response asbeing neuromuscular and the source of resting muscle tone, morerecent work indicates that the source of MV is at least in partcardiovascular (Brumlik, 1962). A lack of suitable instrumenta­tion has obscured both'the source and the nature of thisphenomenon.

The term, microvibration, was first applied to the minutevibratory movement of the human body surface by Rohracher in1946, although he first observed the phenomenon in 1943 (1949,1962b). Other workers studying MV include Denier (1956,1957), Haider & Lindsley (1964), Heller-Jahnl (1959), Marko(1959), Schrocksnadel et al (1956), Sugano (1957), Sugano &Inanaga (1960), and Williams (1963, I964a, b). AlthoughRohracher described the phenomenon as a normal physiologicalfunction involved in generation of resting muscle tone and bodytemperature regulation, Sugano & Inanaga (1960) described it asa kind of minor tremor.

The results of tremor research are closely related to study ofthe human microvibration; the same instrumentation can be usedto study both. Of particular interest in tremor research is thework of Chase et al (1965), Cooper et al (1957), Halliday &Redfearn (1956). Hamoen (1958). Jasper & Andrews (1938),Lansing (1957), Lindsley (1935), Lippold et al (1959), Marshall& Walsh (1956), and Travis (1929). An additional group ofworkers has studied the "normal or physiologic tremor," whichmay be an identical phenomenon. This group includes Brumlik(1962), Bushes et al (1960), Wachs et al (1960), and Wachs&Bushes (196 I). But little transfer of instrumentation techniques

60

has occurred between those studying "microvibration" and thosestudying the "normal tremor."

Significant contributions have been made to instrwnen~tion

by: BekCsy (1941), Bricout & Boisvert (1950), Drischel &; Lange(1956), Hannah et al (1954), Hull (1946), Jackson (1954),Mashima (1956), McAdam &; Moreland (1964), Vinycomb(1954), and Zaalberg.van Zelst (1947, 1948), althoush theseworkers were not specificaDy concerned with MY or tremormeasurements.

Without doubt, Rohrac:her is the pioneer in themicl'O,ibrationfield (1946, 1949, 1952, 1954, 1955, 1956, 1"95Sa, 1958b,1959a, 1959b, 1960, 19618, 1962b, 19648, 1964b). Since 1943,he has tested over 900 Ss, utilizing a variety of instrumentationtechniques, arid has reported the following characteristics ofhuman MV: The phenomenon is always present from birth toshortly after death, and is found in all parts of the body.Observed on the left dorsal forearm, MV has an amplitude of .5to 3.0 microns (1 micron = 39.4 millionths of an in.) withfrequency content predominately between 8 and. 12 cps.Frequency is not the same at all body sites and is lower for largerbut higher for smaller muscles. Smaller animals have a higher MYfrequency and severing the medulla oblongata or removal of theheart does not eliminate the phenomenon. Microvibrationsincrease after exercise, but muscle relaxants reduce theamplitude. Central nervous system stimulants increase MVamplitude, but CNS sedatives reduce it, as do "ganglionblockers." Finally, Rohracher observed that MV frequency isincreased in fever, decreased in a hot environment, and increasedin the cold.

Based upon these observations, and the supporting work ofHeller-Jahnl(1959), Marko(1959), Sugano(1957), and Sugano &Inanaga (l960), Rohracher (1959a, b) concluded that the humanmicrovibration was due to continuously occurring, minute musclecontractions, and he proposed that human MVwas the SOurce ofthe ever present muscle tonus of the striated muscle tissues.Further, Rohracher formulated a hypothesis that the phenome­non was a controlling factor in body temperature regulation(1964b). Observations by Heller-Jahnl (1959), Marko (1959), andRohracher (1955, 1959a, b, 1964b) on MVfrequency confirmedthis theory, but data on MVamplitudes did not.

Sugano(1957) and Sugano & Inanaga (1960) viewedthe MVasa type of "minor" or "invisible" tremor. Rohracher did not holda fundamental objection to the minor tremor view, but statedthat "microvibration" was a better name for the phenomenondue to its continuous presence in normal Ss. He viewed the MVnot as a reaction, but as a biological action of the organism(1949, 1962a, 1962b, 1964a).

Williams (1963, I964b) contended that human microvibration,as measured by Rohracher, was primarily an instrumentationartifact, and that the mechanical constants (mass, spring constant,and coefficient of damping) of an electrodynamic probe(Rohracher used a Phillips PR9260) could act in combinationwith the properties of body tissue to modify the probe response.The modified response results in a probe resonance in the 8 to 12cps region and, thus, generates the apparent MV. AlthoughWilliams disagreed with Rohracher's view of the phenomenon, he

Behav.Res. Meth. & Instru. 1968, Vol. 1 (2)

Including the sensor initial capacitance in parallel with theoscillatorcapacitance, this becomes,

Where, A is the plate area, d is the spacing, and e is the dielectricconstant of air. Fringing fields were ignored. and e was assumedto be €o =8.85 X 10-1 2 farads/meter. Further. it was simplyassumed that the frequency of the oscillator was given by.

where Land C are the total circuit inductance and capacitance.Using these two equations it is possible to take the partialderivative of C with respect to d, and of f with respect to C. Usingthese partial derivatives, the system sensitivity may be expressedin terms of total differentialsas,

(4)

(3)

(2)

(I)eA

-d-C

f = I21TVLC

Af f ( r )Ad =~d r + Co

where, Co is the oscillator capacitance and C is the probecapacitance. The constructed system was operated at 100 mHz(MC) with d =1O.3m, Co = 19/-l/-lfd, and C =l/-l/-lfd. The predictedsensitivity,Af/Ad was then 2.5 x 109 cps/meter, which for ~d =1micron, would indicate an expected frequency change of 2.5kc/micron, With a typical FM discriminator sensitivity of about20 kc/volt, an overall measurementsensitivity of 125 mY/micronwould result.

In preliminary experiments, a number of MY recordings wereobtained from the left dorsal forearm of male Ss. To obtain themeasurements with a I or 2 mm spacing, it was necessary to shavethe area near the probe.

The above derivation of expected probe sensitivity, indicatesthat overall sensitivities as large as 0.125 V/micron can result.This sensitivity appears to have been achieved,but a great deal ofdifficulty was experienced due to operation at 100 Me. At this

distance between S's arm and the condenser plates due todifferences between Ss in ann sizes.

In order to measure the ~tV with a non-contacting sensor, weconstructed a simple transistor feedback oscillator with anexternal capacitance probe connected in parallel with the L-C,frequency determining circuit. The probe consisted of a flat silverplate with an area of 4.15 em", positioned I or 2 mm from thebody surface. The micron amplitude changes in spacing betweenthe body surface and the probe had an associated small changeinprobe-to-S capacitance, which in turn, modulated the oscillatorfrequency. Some amplitude modulation also occurred, but thiswaseliminated in an FM receiver, used as a demodulator.

In order to estimate the sensitivity of the probe it wasassumedthat the probe-to-S capacitance was described by the parallelplate capacitor equation,

Non·Contacting Capacitance SensorsCremer (r907) proposed a capacitance technique to measure

physiological motion. The work was concerned with measure­ment of gross displacements (e.g., the movement of the heart),but was significant in that it constituted the first description of anon-contacting capacitance sensor in physiological instrumenta­tion. Cremer's technique was elaborated by Atzler & Lehman(1932) and Atzler (1935), who used high frequency signals(100·150 MC) to record thoracic dielectric changes related tocardiac activity. Fenning (1937) used a capacitance technique torecord uterine and intrauterine movements, and Bekesy (1941)measured the amplitude of vibration of the auditory ossicles ofthe middle ear usinga shielded capacitance probe.

Capacitive measurement of physiological motion was, thus,well established when Marko used the technique to measuremicrovibration (Rohracher, 1949, 1959a). Marko placed theentire hand or forearm of S into the air-gap of a large capacitorconnected so as to modulate the signal of a 100 kHz oscillator.But Rohracher (1964a), in a reply to Williams, stated that thecapacitive method was unsuitable for comparative amplitudemeasurements because it was impossible to standardize the

INSTRUMENTATION FORMICROVIBRATION MEASUREMENTS

In the studies described in this report, five differentmeasurement techniques were experimentally evaluated. Theseincluded:

A Non-contactingCapacitance Sensor,PhillipsSeismic (Electrodynamic)Sensor,Absolute Linear-Variable·Differential·Transformer (LVOT)

Sensor,Relative LVDTSensor,Accelerometer Sensor.

It should be noted that the accelerometer sensormeasures themicro-acceleration in microns/sec", while the LVDT and thecapacitancetypes measure the microdisplacement in microns.ThePhillipssensormeasures the microvelocity in microns/sec.

tacitly accepted its existence and, in fact, utilized the MV inpsychological studies(1964a).

Brumlik (1962) and Wachs et aI (1960) studied the "normaltremor" in relationship to Parkinsonism; their accelerometertechniques can be shown to be superior to all previousinstrumentation applications in both the MV and normal tremorfields. Brumlik noted that tremor, at rest, appeared similar inamplitude, frequency, and waveform to the ballistocardiogram,and that the normal tremor could not have a muscular originbecause of its persistence after complete neuromuscular"blockade" with succinylcholine. Tremor on intention, on theother hand, might include both ballistocardiographic andneuromuscular components.

Finally, Jasper & Andrews(1938) suggested that a relationshipmight exist between tremor and the EEG waveform, butLindqvist (1941) studied the problem and concluded that therewas none. Later, Kennedy (1959) postulated that the EEG alpharhythm might be generated by vibrations of the conductivecerebral tissue and might, therefore, be an artifact. But Oswald(1961) and Rosner (1961) appear to have provided evidence indisproof of this hypothesis. Apparently, there is no meaningfulrelationshipbetweeneither the EEG and tremor, or between EEGand microvibrations.

Behav. Res. Meth. & Instru., 1968, Vol. 1(2) 61

high fiequency, movement of lOyobjectwithin IeYen1 metersofthe probe can be obIemd U 10 Irtif'act.A number·ofground.IqIOd IIIieIcIiq techniques were attempted, with little aaccea.Miaoribntion recordinp were obtained but, due to these1rtif'1CtI, calibration was not posIible. TheIe difrlCUltiel do not,however, rule out UJe of such a probe. The theoreticalseDlitivityachie¥ed ii, in fact, excessive. It iI felt that by reducing probefrequency to 10.7 MC/Iee (lOd ~rhapl to u low u 100 kc),artifacts due to nearby R.F. field' effec:ts can be eliminated.Equation (4) indicates that reducin& the frequency willproportionally reduce the sensitivity for the lime spacing IOdcircuit capacitance.Fp 1 presents the IChematic of the 107.5 MC capacitIDce

senIOr. The device is basica1ly 10 FM modulated OICiIlator. Thetuned relODlOt circuit, consisting of L

l• C

2, IOd C

3• establiahel

the OICiDator frequency.

Flpre 2 pments mn surface diaplacement recordinp(uncalibrated) obtained wit:h ·the capacitance probe. To obtainthe recordinp, the S wu prone IOd the probe wu poIitionedapproximately 1 mm &om a IhaYed area of the left cIona1mid.foreum. The S reated upon a paddedmedical examinationtable, IOd the entire ann, abducted 45 .. to the micl-aBlaryline, was rested upon a pillow. The reeordinp of FiB. 2cleu1yIbow cudiovucuIar (A) and mpiratory (8) mn surfacecIiaplacements. The recording wu obtained by telemeterina thePM OICiDator IipaI approximately 10 to 20 ft to 10 PMreceim'.

FJectrody--..: ......Electrodynamic vibration tranlllucen may be organized into

two general cafeBories clependina upon how the vibration isapplied to the tranIducer. The fm catepy, pelbapl the mostcommon, inch1des aD Illilmic andabsolUte transducers in which

r-----

al44 &10

-.-,

IJ..T

Fia- 1. MicroribratioD(MY) capedtaDce IeDIOr

ICbematic.

IN7II~

..J

IN1S"­...J

JII

~IL--- J

-----,

L~: : n g~ ;1,: I: ; ; 11 m" !Ii: d 'illl" ;1;1 I'i !I!:la:: : j I 1:' J. lliiijli I !iii L I: i :::' : I ~ iJ ::! ~,; ~ Ii 'ii, i ~ I ' ';1, 1:,:

" ::! ::! :;i::1: ::1: .. .. it ,I; ... .. . ' .. .. ,. I

l ,:i :1, ' =!;: ;!:i 11.; ~ 1: : :i~; t~ : ti ',1' ,i!i !: :: .. :! I !i;; i'ii Illi ' , i j!, I .I' ~ ~ !'I II: :;: ." I! :!; ;,Ii ;!j ~ u.: :;l; t ; t ,i! : ; .- .. .. i 11: i 1 ::;1 ! , i 'j;: ·:t· . .. .. .. ...

iI i, ! ~\ .r:~ ,f h , I;: Ii. ,. r1:Ii •.= , .".l;loil' :1: !:~ I

,Iill. :i; .. .. .~ :

.. .. ..!iii...

,~ :: ~ IIf' .. , ,~ ; I II' I. ,. .. .: : I ..

/~,, '1 , , .. : IH?" II : ... .. ~

t! :! ;;1 ;1:: i':lilOO lili Iii!! ;j: :1: ' , ,1::: :I: :1 iH:: .!h i • :" i I:::;'1 ;j! 1 in: !I!! J:Ii', .: :: .. iii' :i~.. 1:' : .... . . ;n .. . ,.,

" ... ... ...:g ~ ~ lI!

.• •1 1m! tii' ;.1. :1;: il;i i!ll , I. J ~, 1m: ~ I~ i111 Ilfii ilL,Ii! :Ii; ~::i ~. j 1di I ~ lilliliii' u I In: litl: :::j il:! i!;1 jHl :1;, iiI: ,i ii: ~!.i i::: .~ : : .. .. .. ... ,. ..... u:

A

I· i i I: I ;111 n J! ," I·;; ... m: :i;: ;':: :g; illi I' ! ll!: ~ if nli ! I I !i II 1:; ,!!.u I I !~ ! ; t . :j:! ii'; .. , I ,,' "r: : .. .. .. .. .. .. r: .. .' ...

iU' or:: :,1: ::: .. ~i ~: .... ':I! iii' I;!I II:. .. , :Iii lili 'Hili! 1.:: t , ::1 ;! I • . ... il;; .. ;1 y . .., I ... : ~ t .. .., I , .. Ui .ll i ~ : ~ . '\ .. .. !J "'" .. .. I , ·1' j: Ii I n.1"J'o-: 1:

~... , It', :j: .. ... : . ~ Iol n ,

! I· i> II: ~ l 1: . :;r 11'1 U',"'!" , v. '. ..ji 'iii .. :l :: ' j' ~ I!i 11. ,

~l:;' :;~::;;.. .. .. .. ... , .1, ... .. : ~ I , :1: , .. ... .. ..

I II !'II I: :I : , , Ii: l..' ::i:Iii; :l;; ij' .. ," it !:': !LII\.. d. ::1 jE !: ~ ; ... I ::... - .. .. .. .. .. . . u ... .. .. ..

~!: :ii: ltu !ill Ii i ". i .:: i~ iii ~ l:~ :1I~ 'ji,I,'i, :1'1 , illIIlUlil .. . I I : I,;ii iH; .. .. .. ..,- 'C : ... .. ... .. .. .., . .. ::.: .. . . , :'1' .. ..

B

Fia. 2. Capacitance probe recordinp showing pulse IOd respiratory ann-surface diaplac:emeDts for the lIIIle S (A.P.), left donalforearm. Top record is cardiovllCUlar, S to 22 cpl, 100 mV/cm, 1 em/sec. Bottom record is respiratory, de to 22 cpa, 200 mV/em,I cm/1IllC.

62 Behav. Res.Meth.cllDltru., 1968, Vol. 1(2)

(a> Seismic Transducers (b) Relative Transducers

m

Fig. 3. Electrodynamic vibration transducers, aeillmic and relative.

(9)

(ll)

(10)

mWrQ=­c

Using this notation, Equation (8) becomes

X2 • Xl

-x;- = I. jQ(Wr .~)W Wr

Equation (II) is plotted in Fig. (4a) for several values of Q, andclearly shows the resonance of the probe at W = wr• Equation(II) may be simplified at very high or very low frequencies: Atlow frequencies, w«wr, or,

where,wr is the resonant frequency,

(5)

Ifa sinusoidal excitation is assumed, that is,

the vibration is applied to the case of the entire unit (Fig. 3a); amoving mass is suspended within the unit. In the case of thePhillips PR9260 sensor used by Rohracher, a voltage coil and adamping coil are rigidly connected and supponed by two springmembranes. The coils move within the field of a permanentmagnet and generate a voltage proportional to velocity. Thesecond group of transducersmay be termed "relative" in that thevibration measurement is made by a probe relative to the caseofthe unit. The case may either be rigidly clamped to a fixedsupport or may be stationary due to its own largemass(Fig. 3b).

Figure 3a shows the first group of transducers. Thedisp1acemebt to be measured, x.. is applied to the case. Themovable mass, m, is attached to the case by a spring withconstant, k, and damping,c.If the forceson massm are summed,we obtain,

(6)~« W r

Wr W(12)

It can be shown that the response of the mass is also sinusoidal,

(7)

In this region,Equation (11) becomes,

IX\IXI I=I~~21 (13)

and that the relative motion, (X2 • Xtl, due to the excitation,Xl' is

At high frequencies, w»wr' or,

(8)X2 • Xl ~W)2

x;- = OW)2 +jW(~)+ ~

Since typical electrical arrangements respond to (X2 • xj ),Equation (8) will completely describe the frequency response ofthe system.The "Qn of this mechanical systemmay be defined inanalogy to the Q of a tuned electricalcircuit,

~» wrWr W

In this region,Equation (II) becomes,

(14)

(15)

Behav. Res. Meth.& Instru., 1968, Vol. I (2) 63

(a) SEISMIC TUIISOUCas (b) RELATIVE TUIISDUCas

10 lOOK

101.00.1

1--11:-11/

1/

V ~

7

.... /J':/

~IV <! "1/i

II

~

10K

u

O.lK0.01

11;1

'\. Qs ~1.0 '"

II /

11/ J')17

I

/I 1/

II 1/

If/0.1 .0

1.0

0.1

0.01

11e;:11

Fig. 4. Electrodynamic transducer frequency response for seismic (left) and relative ,(right types.

and at low frequencies, W«W r. to be,

which may be simplified at high frequencies, w»wr' to be,

_E- =j (~) ~ (I + jQ (~ . Wr)) (16)

X 2 wr Q \' W r W

reluctance pick-up) can measure velocity only below resonance.For comparison, the Gulton accelerometer to be discussed latermeasures acceleration below resonance.

Figure S shows a Phillips PR9260 seismic velocity probepresumably identical to that used by Rohracher in the major partof this work. This probe was used to measure the microvibrationpresent on the left dorsal forearm of a male S. In all cases, the Swas lying in the prone position. The sensor was supported by awire harness from a heavy frame, and essentially the entire weightof the probe (600 g) was allowed to rest on S's arm. The voltageoutput of the probe was directly recorded on either a BeckmanElectronic Instruments Division R or RS Dynograph Recorder.The S rested upon a padded medical examination table, and theentire arm (abducted 4S deg to the mid-axillary line) was restedupon a pillow. A simultaneous electrocardiogram was taken usingBeckman silver-silver chloride electrodes in lead I, with a sternalground. .

Figure 6 shows resting microvibration measurements obtainedwith the Phillips seismic probe. Figure 6c shows the probe noisebackground without S; Fig. 6a and 6b show the effect ofrestricting amplifier bandwidth to S to 32 cps, as was done byRohracher. Figure 6a was recorded wide-band from de to 150cps. The smaller amplitudes in Fig. 6b show that significantfrequency components do exist outside the 5 to 32 cps band. Todemonstrate the effects of arm muscle tension, a recording wasmade with S's arm maximally tense, but as stationary as possible.Figure 7 shows the effects of tension and includes a simultaneousECG: electromyographic artifacts may be seen in the ECG due tothe arm tension.

The PR9260 measures velocity in the range S cps to 100 cps.The basic probe sensitivity is in mV per em/sec. To measure themicrodisplacement. it is necessary to integrate the probe signal.An accessory, the Phillips PR92S2 vibration meter, can performthe necessary integration for sinusoidal vibrations.

Since Rohracher described his apparatus as consisting only ofprobe, amplifier. and recorder, it must be concluded that MV

(18)

(17)

Equation (18) indicates that the transducer appears to be purely aspring at low frequencies. If the transducer produced a voltageproportional to velocity. X2, then it would measure velocity atfrequencies below resonance.

After examination of both groups of transducers, a majordifference is obvious. The Phillips PR9260 seismic electrody­namic transducer can measure velocity only above resonance. Therelative electrodynamic transducer (for example, a variable

This indicates that X2 becomes zero above resonance, that is, themass becomes stationary. The Phillips probe produces an outputproportional to the first derivative of (X2 • XIl, and will,therefore, measure velocity at frequencies above resonance, andthe rate-of-change-of-acceleration (rarely used) at low frequen­cies. Even if critically damped, this probe will be, at best, difficultto use near resonance.

The second group of transducers is best analyzed by solvingforthe displacement X2 resulting from an applied force, F. Figure(4b) illustrates this system. Writing the summation of forces, asabove, it can be shown that the sensor response is,

64 Behav. Res. Meth. & lnstru., 1968, Vol. 1 (2)

,, I, I

- , I~ .. i I

.~l" r.;

~

"

-t r,

i .". ~. : ;. ~ '....

'-1' .i : '. .i. .1.. :1. - .

! -. - -

-, I

.IT- - -- - - j-

A

LL

:..

B

C

Fig. 6. Resting MY recordings obtained with seismic sensor.(A) MV: de to 150 cps, 2S mm/sec, I mV/cm. ECG: Lead I(sternal ground), .5 mY/em, Ag/AgCI electrodes. S: M.T., 28 yr.old male, left dorsal forearm. (B) MB: S to 32 cps, 25 mm/sec,1 mY/em. ECG: lead I (sternal ground), .5 mY/em, Ag/AgCIelectrodes. S: M. T., 28 yr old male, left dorsal forearm. (C)Probe noise without S, I mY/em, 25 mm/sec, 5 to 32 cps.

Unear·Variable-Differential Transformer Sensors (LVDT)To compare the electrodynamic and noncontacting capacitance

measurements with a measurement made by a non-resonant,contacting probe, several sensors employing linear differentialtransformers were designed and constructed. Sanborn"Linearsyn" LVDT transducers were used with core weights of18.0 g and i l g.

An LVDT sensor may be considered to be a type of absoluteprobe in that it measures with respect to a fixed mechanicalreference. The mechanical reference may be independent of S orthe sensor may mount on S and measure displacement of oneportion of the body relative to another. A simple 2.4 kc/seccarrier system was designed for use with these sensors and theentire system was constructed as a plug-in coupler compatiblewith a Beckman Electronic Instruments Division type R or RSpynograph Recorder.

velocity was measured, and that calibration was in terms ofdisplacement. It is true that the integral of a sinusoidal velocity issimply a sinusoidal displacement with shifted phase, but it maynot be assumed that the MV is purely sinusoidal. The recordingsabove clearly show this.

Halliday & Recrearn (1956) have pointed out that it is notsufficient to report tremor as being of one particular frequencyunless it is made clear that either displacement or its flrst orsecond derivative is being discussed. The authors also point outthat electrodynamic systems can measure velocity while opticaltechniques measure displacement. These comments, althoughgenerated in regard to instrumentation for tremor measurement,apply equally well to measurement of the MV. This would tendto explain the differences in amplitude noted by Marko (1959)when using an optical technique, and Rohracher, using anelectrodynamic technique. Although the PR9260 is damped forsinusoidal vibrations, the probe exhibits a strong tendency toresonate in the frequency range of interest. Although frequentlyused, this probe does not appear to be suitable for MVmeasurements due to its resonance near 10 cps, and its largephase error (approximately 90 deg at 10 cps). Figure 8 presents agraphic demonstration of the resonance problem. The Phillipsprobe was positioned at (A) a specific blood vessel and (B) at thedorsal mid forearm, without specific attempt to locate a bloodvessel. In both cases, however, it was observed that a vibration ofvascular origin, at the arm, apparently excited the probe, andgenerated a burst of 10 cps "false microvibration" signal. Thedelay time observed between the ECG R wave and the onset ofthe false microvibration signal is comparable to the pulse-wave­transit time found in pulse-wave-velocity studies.

To further demonstrate the resonance problems associatedwith the Phillips seismic probe, recordings were obtained aftertapping the probe. Such an excitation is, of course,non-quantitative and much larger than microvibration, butnevertheless serves to exhibit the probe resonance. Figure 9 showsrecordings obtained with the probe suspended (no S) and tappedin its axial direction. Sustained oscillations may be observed afterthe excitation.

Fig. 5. Phillips PR9260 seismic velocity probe.

Behav. Res. Meth. & Instru., 1968, Vol. 1 (2) 65

Fig. 7. Microvibration during arm muscle tension with seismic sensor on 28 yr old male S (M.T.), left dorsal forearm. Upper recordshows MV, dc to ISO cps, 20 mV/cm, 2S mm/sec. Lower record indicates simultaneous ECG with electromyographic artifacts, lead I(sternal ground), .SmV/cm.

A

8

Fig. 8. Microvibration in relation to pulse with the seismic sensor used on same S as in Fig. 7. Upper recording (A) shows probepositioned at a specific blood vessel. Lower record: Same probe positioned at dorsal mid-forearm. no SDecific attemet to locate bloodvessel (A) MV: 5 to 32 cps, 1 mV/cm, 25 mm/see. ECG: Lead I (sternal,ground),.5 mY/em, AglAg£1electrodes. S: M.T., 28 yr oldmale, lett dorsal forearm over cubital vein. (B) MV: dc to 22 cps,S mY/em, 25 mm/sec. ECG: Lead I (sternal ground), .5 mY/em,AglAr£-1 electrodes. S: M.T., 28 yr old male, left dorsal forearm.

Although the sensors were initially constructed withDelrin-Delrin bearings, it was found that lubricated metal bearingshad lower friction forces. Calibration of the LVDT rnicrovibra­tion sensors was performed, using precision gauge blocks with

66

heights known to within +2, -4p. in., using an optically flatreference block. Sensitivities of 58.4 p.V/p. and 66 p.V/p. wereachieved for two sensors using the small and large core LVDTs,respectively,

Behav. Res. Meth. & Instru., 1968, Vol. 1 (2)

Fig. 9 . Resonant characteristics of Phillip's seismic probe, suspended, with no S, and tapped in axial direction (dc to ISO cps,25 mm/sec, 5 mY/em).

Fig. 10. Relative S-mounted LVDT sensor (top), and absoluteframe-mounted LVTD (bottom).

Behav. Res. Meth. & Instru., 1968, Yol. I (2)

11 1H1lm:1l1liJli Uli llilll~ 1 III1111!!l11111 11'}IIIIIB ISUUIDI~ I I I :I I IW>llIIllm "l lm,\ . '

""' -

Fig. II . Accelerometer sensor, Gulton AVR·2s0-30.

Preliminary S tests with the arm-mounted "relative" LVDTsensor, shown in Fig. 10, revealed an absence of MV signals.

Microvibration signals were observed, however , with the sameLVDT sensor and S, when the measurement was obtained withrespect to a fixed frame. It was concluded that the human MVdisplacement is not a local phenomenon. It was possible to obtainmeasurements with the frame mounted LVDT sensor , but th isapproach was not considered useful because of an obviouslimitation. Use of an absolute frame-mounted LVDT sensor

67

rr

­,

,."'~T""

I VItI. '"!>' I ~ft!.' 7. ) ..._..I fI':" 0 "

,.... ~~ I ~" ~~.M.... e.' I

II

«J~I ~ .!1t

-e- I " c.ooc.o& -"'0..~: .~~. 1 ·'$ 1(.l ~lPn '--

, ..."",.'" ......'"

~ia. 12. MYaccelerometer carrier system, schematic diagram. All transistors 2N3565; L1 isTRIAD EM.202, 20 mh; TI is TRIAD SP4.

+i i r-rT r l, I I I i rrTrl i '.1A

~ig. 13. Microvibration recordings with accelerometer sensor on same S(H). MY: 5 mY/em, ECG: .5 mY/em, 25 mm/sec.

requires the S to maintain the body-to-sensor distance within afew thousandths of an inch, so as not to exceed the sensordynamic displacement range.

Accelerometer SensorsThe accelerometer shown in Fig. II (Gulton AVR-250-30) is

basically a variable reluctance device and includes two activeinductive elements. The sensor was first reported applied intremor research by Brumlik (1962), Wachset al (l960), Wachs &Bushes (1961), and Bushes et al (l960). A seismically suspendedmagnetic armature varies the air gap of each inductive arm inproportion to acceleration. Since the magnetic armature willremain displaced in response to a fixed acceleration, the sensorwill respond to acceleration frequencies as low as zero cps. Byrotating the sensor 90 deg, the acceleration of gravity may beused as a convenient calibration.

This sensor responds only to accelerations perpendicular to theplane of the case. Since the sensor resonance is at 250 cps, theuseful frequency range extends from zero to over 100 cps. Thetotal weight is only 2.5 g and overall dimensions 7/8 x 17/32 x3/16 in.

The schematic for the associated carrier system is shown in Fig.12. The accelerometer carrier system is similar to the LVDTsystem mentioned earlier except that the sensor is operated in anelectrical bridge and a carrier amplifier is provided. The sensordriving signal is obtained at the 2 KC oscillator (QI).'TransistorsQ2 and Q3 form an isolation or buffer amplifier. The sensorbridge includes resistive and inductive balance controls so that atrue bridge null may be obtained. The operational amplifierprovides a gain of 200 to amplify the bridge unbalance signals.The carrier demodulator is a standard voltage doubler. A filter isprovided at the output to reduce the carrier frequencies. The

+r+f++-++4-!.J .J...J.+.:. ! '. ' ' ,

-'>.

-t

I ,, , ,1. 1

.~I

!!

.t - ~ ..l~

I I

.. - .-

1'fl I l ~ v'.\ rL ,~ftH I I r;~:rr+4: ~

Fig. 14. Simultaneous phonocardiogram and MY (microacceleration) recording from chest of 26 yr old male (D.R.). MY: (Upperrecording) 20 mY/em, noise 2-3 mY, 2.5 em/sec. Phoncardiogram: (Lower recording) 5 mY/em, noise 2 mY.

68 Behav.Res. Meth. & Instru., 1968, Yol. 1 (2)

overall system sensitivity was 1.1 Vjg of accelerat ion. Figure 13illustrates typical experimental records obtained with this sensor.The simultaneous ECG was transthoracic with Agj AgCIelectrodes and a sternal ground. Prominent cardiovascularcomponents are obvious in both records.

With regard to the cardiovascular components in the humanmicrovibration, simultaneous accelerometer-Mv and phonocar­diograph measurements were made on the chest (Precordium).The phonocardiograph designed for the NASA Gemini programwas used. Very strong accelerations were observed on the chestsurface, bearing a direct correlation to the phonocardiogram andelectrocardiogram. Figure 14 shows a simultaneous rnicroaccelera­tion (upper) and phonocardiogram (lower). The first and secondheart sounds visible in the phonocardiogram are clearly seen inthe MV recording. These preliminary observations tend toconfirm the existence of a large cardiovascular component in thehuman microacceleration.

CONCLUSIONSThe existence of human body surface microvibratiors was

verified; some previously used instrumentation techniques werefound inadequate. In preliminary experiments, it was observedthat the body surface microacceleration apparently containssignificant cardiovascular components. This result is in oppositionto the muscular and temperature regulation theory of Rohracher(1946-1964).

Accelerometer sensors proved to be most satisfactory formeasurement of the microacceleration; several types ofcommercial units are available. Due to small size and mass, suchsensors can be mounted by means of doublefaced adhesive tapeto almost any portion of the body, yet allow relative S freedom.Measurement of acceleration is experimentally superior, as a fixedsensor reference is not needed. The capacitance sensor, althoughoffering the advantage of not contacting the body, was extremelysensitive to artifacts from gross motions.

The Phillips selsmicprobe was much too large and massive, andalthough critically damped, exhibited resonance in the microac­celeration spectrum at 10 cps. When excited, this transducerproduced a 10 cps signal not unlike the microvibration itself.Near 10 cps, phase errors also i>resented a difficult problem.Seismic probes can be shown to have a low frequency resonancethat severely limits their usefulness in measurement of the MV.

The LVDT sensors experienced bearing friction problems. Asmall frame-mounted (absolute) LVDT sensor was attempted butrequired the S.to maintain body-to-sensor distance within a fewthousandths of an inch.,an obviously impractical-requirement.

R£FERENCESATZLER, E. Dielektrographie. In: Handbuch der Biologischen Arbeits­

methoden; Berlin-""Wien: Urban and~chwarzenbeJg, 1935. Vol. 5. Pp.I(173-1084.

ATZLER, E., '" LEflMAN, G. Ober ciA neues VClUhrea zur Darstellung.de~'Herztatigkeit.Arbeitsphysiologie, 191~, 5, 636-680.

8EKESY, G. von. Uber die Messung der Schwingungsamplitude derGehorknochelchen mittels einer kapazitiven Sonde. Akustiche Zeit­schrift, 1941, Jan., 1-16.

BRICOUT, P. A., & BOiSVERT, M. Measurement and amplification ofminute displacements by frequency modulation. Rev. Sci. Instr., 1950,21,98-99.

BRUMLlK, J. On the nature of normal tremor. Neurology, 1962, 12,159·179.

BUSHES, B., WACHS, H., BRUMLlK, J., MlER, M., & PETROV1CK, M.Studies of tone, tremor, and speech in normal persons and Parkinsonianpatients. Neurology, 1960, 10, 805.

Behav. Res. Meth. & Instru., 1968, Vol. I (2)

CIIASE, R. A., CULU.N, J. K.. & SUI.LlVAN. S. A. Modification ofintention tremor in man. Nature, 1965, 206,485-487.

COOPER, J. D., HALLIDA Y, A. M., & RI DFEARN, J. W. T. Apparatusfor the study of human tremor and stretch reflexes. t't'(; clin.Neurophysiol.. 1957,9,546·550.

CREMER, M. The recording of mechanical processes by electrical means,particularly with the aid of the string galvanometer and stringelectrometer. (in German). Muenchener Medizinische wochenscbriit,1907,33,1629'1630.

DENIER, A. Les microvibrations du corps, tremoins du tonusphysiologique, Revue Neurologique, 1956,95,503-506.

DENIER, A. The microvibrations of the body as an expression ofphysiological tonus. t'EG clin, Neurophysiol., 1957,9,362·363.

DRISCHEL, H., & LANGE, C. Uber unwillkurliche Augapfelbewegungenbei einaugigem Fixieren. PflugersArchiv.. 1956,262,307·333.

FENN1NG, C. A new method of recording physiological activities,l and lI.J.lab. clin. Med.. 1937,22,1279-1284.

HAlDER, M., & LINDSLEY, D. B. Microvibrations in man and dolphin.Science, 1964,146,1181-1183.

HALLIDAY, A. M., & REDFEARN. J. W. T. An analysis of thefrequencies of finger tremor in healthy subjects. J. Physiol.. 1956, 134,600-611.

HAMOEN, A. M. On the physiology of tremor. t'EG clln, Neurophysiol.,1958,10,752·753.

HANNAH, K. W., JONClCH, M. J., & HACKERMAN, N. An automaticsystem for the study of catalytic reactions involving gases. ReI'. Sci.Instr., 1954,25,636-639.

HELLER·JAHNL, I. Die Mikrovibration bei psychischer Spannung undEntspannung. Zeitschrift Psychother, Med, Psychol., 1959,9,34·38.

HULL, G. F. Resonant circuit modulator for broad band acousticmeasurements. J. appl. Physics, 1946,17,1066·1073.

JACKSON, P. Resistance strain gauges and vibration measurement. J.British IRA:, 1954,14,106·114.

JASPER, M. M., & ANDREWS, H. L. Brain potentials and voluntarymuscle activity in man. J. Neurophyslol., 1938, 1, 87·100.

KENNEDY, J. L. A possible artifact in electroencephalography. Psychol.Rev.. 1959,66, 347·352.

LANSING, R. W. Relation of brain and tremor rhythms to visual reactiontime. Et'G clin, Neurophysiol.. 1957,9,497-504.

LINDQV1ST, T. Finger tremor and alpha waves of the electroencephalo­gram. Acta Medica Scandinavica, 1941,108,580·585.

LINDSLEY, D. B. Electrical activity of human motor units duringvoluntary contraction. Amer. 1. Physiol... 1935, 114, 9()"99.

LIPPOLD, 0. C. J., REDFEARN, J. W. T., & VUCO, J. The influence ofafferent and descending pathways on the rhythmical and arrhythmicalcomponents of muscular activity in man and the anaesthetized cat. J.Physiol., 1959, 146, 1·9.

MARKO, A. R. Optisch-rnikroskopische Registrierung der Mikrovibrationdes menschlichen Korpers,Mikroskopie, 1959,14, 102·l{)5.

.MARSHALL, J., & WALSH, E. G. Physiological tremor. J. Neural.Neurosurg; Psychiat., 1956, 19, 26()"267.

MASH1MA, H. Recording technique of mechanical changes. Rec, adv, Res.nerv, System, 1956,1,179·191.

McADAM, w., & MORELAND, G. R. A magnetic nuu-balancedisplacement measuring system ~sing~adio·frequency modulation. IEEETrans. Instr. Meas., 1964, 1M-l3, 94-102.

OSWALD, I. On the origin of the EEG alpha rhythms. Psychol. Rev•• 1961,68,360-362.

ROHRACHER, H. Schwingungen im .ruenschlichen Organismus. Anz.Phil.-Hist. Ost. Akad, Wiss., 1946, 23()..Z45.

ROHRACHER, H. Mechanische Mikraschwingungen des menschlichenKarpen. Wien: Urban and Schwarzenberg, -1949.

ROHRACHER, -H, Neue Untersuchungen uber biologische Mikro­schwingungen.Anx. Phil.·Hist. Ost, Akod. Wiss., 1952,153·161.

ROHRACHER, H. Neue Untersuchungen uber biologische Mikro­scbwinlwll2eo 11. Anz. Phil.-ffist. Ost. Akad. -WiGs.. 1954. 229-248.

ROHRACHER, II. Warmebaushalt und Mikrovibration. Acta Neurovegeta­ltVa, 4'J'~, 11, Ill/·lUU.

ROHRACHER, H. Uber eine standig vorhandene mechanische Mikro­schwin,gung des menschlichen Korpers. Ber: ge«: Physiol., 1956, J.80,119.

ROHRACHER, 1i. Neue -Untersuchungen tiber die Mikrovibration desmenschlic1ten Korpers. Anz. Phil-Hist. Ost. Akad, Wiss.. 1958a, 59·78.

ROHRACHER, H. Muscular micro-activity ("microvibration'') as anindicator of physiological tension. Percept. mot. Skills. 1958b, 8, 150.

ROHRACHER, H. Continual muscular activity ("micIovibration") tonusand constancy of the body temperature. Zeitschrift fur Biologie, 1959a,38-53.

ROHRACIIER, II. Untersuchungen uber die Warmeproduktion ZUI

Konstanthaltung der Korpertemperatur. AIlZ. Phil-Hist. Ost, Akad, Wiss.,1959b, 229·240.

69

ROHllACHER, H. Methoden zur R~ IIIld ADnoortuJII derMiboribration. Psycholo&Udte Btl".. 1960.4.118-126.

ROHRACHER. H. The bioIoP:aI function of the lIIpharhythm. EEGclbl.NeruopIIyrilll., 1962a,14, 791.

ROHRACHER, H. Pennanente rhytlunilche MDao~ delWarmbilltenJlpnirmus (''mikroYibntion''). Die NlltlIrwi~1962b. XUX, 145-150.

ROHRACHER. H. Comments on • relOnance theory of miaoribntioDLPrychol. Rn., 1964., 71. 524-525.

ROHRAC¥ER, H. Microvibntion, pam_t mU1C1e ectivity andcOllltanCyof body temperature.Perupr. mot.Skill" 1964b, 19,198.

ROSNER, B. S. Alpha rh)1hm of the EEG and mecbaDk:l1 pIOputieaofbnin.Pryclloi. Rn., 1961,68,359-360.

SOfROCKSNADEL, H., UENER, A., .l SEMENITZ, E. Zur biokJIia:benWirtlllII mlbomcchanbcher SchwiJlpnpn yon IlItnICballfrequenz.Ber.;s. Pfryliol., 1956, IBO. 120.

SUGANO, H. Studies on the microvibration. KUll.lme Mediclll J., 1957,4,97-113.

SUGANO, H., .l INANAGA, K. Studies on minor tremor. JfI1Hl1I J.Phyliol., 1960,10,246-257.

TRAVIS, L. E. The relation of voluntuy movement to tremors.J. expoPsychol.. 1929,12,515-524.

VlNYCOMB, R. K. Electronic aids to vibration me_nt. J. BririlhIRE., 1954,14,115-127.

WACHS, H., BRUMLIK, J., .l BUSHES, B. Studies in PartiDaoniJm:Relationlhips between toDe and tremor in ParIdnaoniJm and in normal..l "trI!. _lit. o«, 1960, l-i.

WACHS, H., .l BUSHES, B. Studiel of tremor In IlO1'IIWs and inPukinlOnUm. Arch Neuro/., 1961,4,66.

W1LUAMS, J. G. L. A relOnance theory of "mkrovibratioDL" Psychol.Re«, 1963,70,547-558.

WILLIAMS, J. G. L. Ute of a reIODaDce tel:hniqlle to meuue mulCleactivity in neurotic and IChizophrenic patients. PrycholDm Med1964a, 24, 2()'28. . .,

WILLIAMS, J. G. L. A resonance theory of "microvibntioDs" a reply toRoIIncher. Pl)'Clloi. Rn., 1964b, 71, 526-527.

ZAALBERG VUl ZELST, J. J. Circuit for condeDlet miaophoDel with lownoile level "'Uipr Tech. Re«, 1947/1948,9,357-363.

NOTE1. The author achnowledps the lIipitlcallt contributions of lib. Merrick

Troupin, Mist Joyce BdiI, and Mr. David Reed. This work woaW havenotbeen pollible withOllt the IIIpportof Dr. Thomas Weber.

The roots of territorial marking In the Mongolian gerbU:A problem of species-common topography

D. D. THIESSEN, UNIVERSlTY OF TEXAS, Austin, Texas78712

The Mongolian gerbil (Meriones unguiculatus) has recentlybeen introduced into behavioral research. Many features make itan ideal laboratory animal. It isdocile, highly exploratory, a goodleamer, Virtually odorless and can be maintained without water;other than that it metabolizes from its food. Generalcharacteristics of the gerbil are described, and a brief review ofbehavioral research is given. The gerbil also possesses uniqueattributes that can only be studied by matching experimentalmethodology with species-common responses. Territorialmarkingis used as example. The gerbil regularly marks objects in an openfield by skimming the object With a midventral sebaceous scentgland. The marking and gland are dimorphic, with the malemarking about twice as frequently as the female and possessingagland roughly twice the size. The configuration of the field(object quality) modifies the frequency of the response, as doesthe time of day the animals are exposed to the field. Androgenlevels control the level of marking in the male and female. and thecorrelations between testis weight, size of the sebaceous gland,secretory output and marking frequency are significant. When agerbil is introduced into an open field recently contaminated byanother gerbil, or when objects are smeared with sebum andplaced in the field, the male tends to be more hesitant in severaltypes of behavior. The laboratory measures are internallyconsistent and congruous with the notion that natural selectionhas acted to reinforce a hormone-behaviar relation of socialsignificanceto the gerbil.

The Mongolian gerbil (Meriones unguiculatus] is rapidlybecoming a significant target for biological research. Most of thework is physiological in character (see Schwentker, 1963), butinterest is also focusing on behavior. In our laboratory we haveconcentrated on three species-common components of gerbilbehavior-spontaneous epileptiform seizures found in a propor-

70

tion of the population (Thiessen, Lindzey, & Friend, 1968),apparent lack of depth perception (Thiessen, Lindzey, Blum,Tucker, & Friend, 1968) and territorial marking with a ventralsebaceous gland (Thiessen, Friend, & Lindzey, 1968; Lindzey,Thiessen,&Tucker, i 968).

The first section of this paper describesvariousfeatures of theMongolian gerbil that will be of value to those interested inmaintainingand breeding the animal in the laboratory•

The second section describes the marking response andillustrates the effects of altering the configurationof the territoryand the circadian phase of testing. The third section gives therelations between marking and other variables and demonstratesthe hormone control of the behavior. It also presents an easytechnique for measuring the pheromone involved in territorialmarking. Last is a discussion of the social and evolutionarysignificance of marking. Attention is focused throughout on howweU the instrumentation of laboratory tactics has elicitedspecies-<:ommon reactions of importance to the adaptation of thegerbil.

CHARACTERISTICS OF THE MONGOLIAN GERBILOassification

The Mongolian gerbil, Meriones unguicuJatus, was firstdescribed by Milne-Edwards (1867). It is a native of northeastChina, eastern Mongolia and Korea. Dr. Victor Schwentker of theWest Foundation introduced the gerbil into the United States in1954. Of the original eleven pairs, five females and four maleswere induced to breed. Progeny of these matings served as thehistorical nucleus for most research colonies now in existence.Victor Schwentker at Tumblebrook Farms in Brant Lake, NewYork, remainsthe major supplierof researchanimals.

According to Schwentker (1963, 1968) gerbils are classifiedinto 10 and possibly 12 genera. The Mongolian species,unguiculatus, is in the genera Meriones, and is further recognizedas part of the subfamily, Gerbillinae, the family, Cricetidae. the

Bebav. Res. Meth. & lDstru., 1968, Vol. 1 (2)