Oct. 8, 1946. 2,409,033 L. GARCEAU ELECTROENCEPHALOGRAPH DEVICE Filled Nov. 4, 1941 7 Sheets-Sheet 1 \NVENTOR Luvs-:1"; .E'RRGEHL! BY l/WM + ATTORNEYS . Oct. 8, ‘1946. ' L. GARCEAU ‘2,409,033 ELECTROENCEPHALOGRAPH DEVICE Filed NOV. ‘4, 1941 8 7 Sheets-Sheet 2 42 60 PR 70 b; INVENTOR. LDVETT EHRCEFILJ A‘ILORNEYS . Oct’ 8, 194,6. ‘ L. GARCEAU ELECTROENCEPHALOGRAPH DEVICE File'd Nov. 4, 1941 2,409,033 _ '7 Sheets-Sheet 4 FIE--5 64 76 70 .75 INVENTOR, LUVETT E'FIR'EEHLJ ATTORN EYS'. Oct. 8, 1946. 2,409,033 L. GARCEAU ELECTROENCEPHALOGRAPH DEVICE Filed Nov. 4, 1941 7 Sheets-Sheet 5 ENVENTOR ' is V; av mATTORNEYS Oct. 8, 1946. |_. GARC'EAU 2,409,033 ' ELECTROENCEPHALOGRAPH DEVICE Fil'ed Nov. 4, 1941' FIEJI. 'T Sheets-Sheet 6 FIEJE QM 4 4W» ATTORNEYS‘. Oct“ 8, 19466 '- I _ 1.. GARCEAU 2,409,033 ELECTROENCEPHALOGRAPH DEVICE Filed Nov. 4,‘ 1941 7 Sheets-Sheet '7 KCHANEL <5 <2 E: _' '2 ‘2 : Q m no r. I w \n I t 3:1] 5, " I ’ 00 m (u — P ' ?g I’ ' ‘i ' 3g é L? ENVENTOR '1 Lever?’ E’HRC'EHLJ L11 LL. ATTORNEYS. 2,409,033 Patented Oct. 8, 1946 UNITED STATES PATENT OFFICE 2,409,033 ELEC'I‘BOENCEPHALOGRAPH DEVICE Lovett Garceau, .Holliston, Mass. Application November 4, 1941, Serial No. 417,865 12 Claims. (Cl. 128—'2.1) 1 2 Referringnow to the drawings: This invention relates to improvements in elec troencephalcgraph devices and has for an object Figure 1 is a perspective view of my new and the provision of a simple and reliable means for recording electroencephalograms and other elec improved electroencephalograph device with the trical electrophysiological potentials. Figure 2 is a front elevation of the device shown in Figure .1, said device having been .re moved from the cabinet; cover opened and partly broken away; Another object of the invention ‘is the provi sion of an electroencephalograph device provided with effective shielding, the provision of attach Figure 3 is a rear view of the device shown in ments for connection to said device and to a Figure 2; shielding devices making it unnecessary for the patient to be placed in a “Faraday cage.” Another object of the invention is the provi sion in apparatus of the character described of a record feed device for advancing the record at a uniform and de?nite rate, and the provision the bottom thereof; Figure 7 isan elevation of the recording tape mechanism removed from the device and show patient, said connections being provided with 10 Figure 6 is a view of the device as seen from ing the portion of the device engaged thereby in elevation; of a stylus traversing a station over which the record passes for recording wave forms or other graphs on the record. A further object of the invention is the pro vision in a thermionic device of an ampli?er in cluding at least two tubes in its input stage, the provision of a multiplicity of electrodes and switching means associated with said electrodes and the input grids of said tubes, said switching devices being arranged so that, for example, two of the electrodes may be selectively associated with said input grids and the remaining elec trodes being all connected together and ground ed,'0r separately grounded through individual re sistive‘paths. Yet another object of the invention is the pro vision in a thermionic device of a plurality of Figures 4 and 5 are, respectively, right and left end views .of the device shownin Figures 2 and 3; Figure 8 is an end elevation of the recording tape mechanism as seen along the line 8—8 of v20 Figure .3; Figure .9 is a plan view of the portion of the device for containing the recording tape mech anism and showing details of the tape drive and therecording mechanism; Figure 10 is a diagram of the internal circuits of the device; Figure 11 is a plan view of one form of elec trode support, the same being adapted to be applied to .the headof the patient for supporting the electrodes in a very large number of combi nations of positions; Figure 12 is a, transverse sectional elevation of the electrode holder shown in Figure 11; Figure 13 is a view showing a shielded cable shielded conductors connected thereto, a plu rality of electrodes for contacting human ?esh 35 carrying a plurality ofccnductors for attach ment to the electrodes shown in the holder in or tissue, a support generally conforming to the Figures 11 and 12; ~ shape of a portion of the human head and adapt Figure .14 is an enlarged view showing details ed to support electrodes in contact with human of one .of the electrodes; and flesh or-tissue in a desired selective arrangement, with respect to each other, and connections be- 40 Figure 15 is a diagram showing theswitching arrangementfor a plurality of electroencephalo tween said electrodes and said shielded conduc graph channels. ' tors. By employing an ampli?er which includes at A further object of the invention is the provi least two input tubes (or two control grids in a sion of a conductorand electrode for-contact with multiple tube), and by providing a plurality of the human flesh, the face of said electrode being electrodes and switching means associated with concave in form for containing a chemical sub the electrodes and said input grids, I am able to stancefor reducing the surface contact resistance selectively associate any two of the electrodes between said electrode and said ?esh or tissue. with said input grids, the remaining electrodes Other objects and advantages of the inven all being connected together and grounded. tion will be apparent to those skilled in the art. 3 2,409,033 In such an arrangement, the special selector switch may have one arm connected to one of the input grids and the other arm to the other input grid. The switches may have ten posi tions for ten electrodes. The switches connect said grids to any two selected electrodes and si multaneously connect the remaining eight elec trodes together and to the ground. _ The switch arms when rotated from one posi trodes will presently be described~the magnitude of the contact resistance through the skin is of the order of 3,000 to 15,000 ohms. This low con tact resistance has the further advantage of per mitting the operation of the ampli?er without appreciable disturbance by the thermal agitation of electrons in the ohmic resistance comprising the input circuit. The grids, in the input stage tion to another short-circuit adjacent contact 10 of the ampli?er, are connected to ground through resistors of the order of 100,000 ohms, thereby points while bridging the spaces therebetween. in effect further limiting the maximum imped The grids are never "free” and cannot receive ance across which an interfering ?eld may build large signal voltages which might temporarily up a voltage. paralyze the ampli?er. A shield is provided between the primary and In order that a clear understanding of the elec the secondary windings of the power transfor trical properties of the human body may be had, mer, and the power cord leading from said pri the following should be borne in mind: If the body were a perfect conductor according to Fara day’s laws, all parts of the surface would be of the same potential, and it would, therefore, be im possible for my new and improved electroenceph alograph device to detect upon it any differences in potential resulting from action currents. Were the body a perfect insulator, any poten tials developing within it could not be led off and recorded by a device having ?nite input imped ance. As a matter of fact, the body is a poor conductor, therefore potentials developing there in at various points can be led off by surface elec trodes to a system having input impedance of 3 O comparable or larger magnitude. Adjacent and more distant portions of the body act as an antenna capacitively coupled to near by conductors, such as power lines, electrical ap paratus, etc, and pick up interfering voltages which are conducted through the body to the lead-off electrodes of the recording system. In order to drain off to ground a substntial portion of the charges induced by such coupling, the electrodes not connected to the system are grounded either separately through a multiplicity of resistive paths or all directly. Interference from electro-static ?elds is further reduced by placing the patient upon a com paratively thin mattress or cushion which rests in turn upon a grounded conducting metal table or a conducting bed-spring. This increases the capacity between the body and earth or ground and therefore reduces the potential induced in the body by any given strength of electro-static mary to a lighting circuit outlet is also shielded. The grounding system is designed so as to avoid loops in which an interfering electro—magnetic ?eld might build up a large circulating current and across parts of which an interfering voltage might therefore develop. I have found that a ground line made as follows is very effective: A ?exible wire ending in a clip and forming a part of the subject’s electrode harness is effective, particu larly when the clip is connected to the metal of the bed spring or to the metal of the examining table. This ?exible wire is also connected to the ampli?er case. From its point of arrival inside the case, the wire runs as directly as possible to the cathode of the ampli?er tubes in the ?rst stage. All other grounds within the ampli?er circuit are made either to this wire or to the case or chassis. The shield of the power-cord may be connected to this wire or to a grounded point within the ampli?er case and at the end where the power-cord plug enters the lighting circuit socket the shield is connected to another ?exible wire carrying a clip which is connected to a wa ter-pipe ground. Further means are also employed in my new and improved apparatus to reduce interference; for example, it will be noted that the cathodes of the two triodes in the first stage are connected directly to ground without any bias resistor. These tubes operate with no grid-bias. By di rectly grounding these cathodes, any potential which might be induced upon them by capacity coupling through the heater insulators from the ?eld. heaters, which are operated on raw alternating As a further step in the reduction of interfer ence, the ampli?er operates on a push-pull prin current, is eliminated. This arrangement also eliminates potentials due to ohmic leakage through imperfections in the heater insulators. In the subsequent stages of the ampli?er, how ciple and therefore the potential differences be tween the grids will be ampli?ed, while the po- .. tentials induced by the above-mentioned capaci tive coupling affects both grids equally and with the same polarity and will not be ampli?ed. In this speci?cation, the differences in poten— tial between the grids will be termed “potentials .; which are out of phase with each other,” “po tentials which are out phase with one another,” or "bio-electric potentials.” Potential changes upon both grids in the same direction or polarity will not be ampli?ed, and these potentials will be termed "potentials in phase with each other,” “potentials in phase with one another,” or “in terfering potentials.” The means by which this is accomplished will be presently described. To reduce the interference pick-up as much as * possible, contacts established with the surface of the tissue are of a minimum feasible resistance, thereby substantially reducing the impedance across which an interfering ?eld may build up a voltage. With the electrodes used—which elec ever, the cathodes are biased by means of bias resistors and capacitative and ohmic leakage is therefore balanced as far as possible, for exam ple by the expedient of grounding a mid~point of a resistor of 100 ohms bridging the heaters. In the plate supply of the ?rst and second stages of the ampli?er, I have introduced a volt age regulator tube which is effective in the re moval of 60 cycle or 120 cycle ripple from the ?lters following the recti?ers, and it is also ef fective in an even more essential capacity in maintaining constant plate voltage for these stages. This is very necessary and important be cause relatively small plate supply voltage changes would be ampli?ed by the system, and, therefore, would appear as interfering signals in the third and fourth stages, and these interfering signals may be of sufficient amplitude to block the ampli?er. The residual 60 cycle interference from power 2,409,033 lines is ?ltered out by a ?lter. choke introduced between the third and fourth stages.» lv have found that it is preferable to place the ?lter in this position in the circuit because here the sig nal strength is large enough not to be over-ridden by the hum picked up by the inductance ele ments from the power supply chokes and the power transformer, and also the hum component of the signal at this point is not large enough to run the ampli?ers oif their linear character istics and thus modulate the action~current sig nal. would be conducive to enormous degeneration effects, but I prefer to use the circuit arrange ments herein described. In the last stage of the ampli?er, the push pull arrangement herein shown is responsive to only those. signals which are out of phase with each other, and signals which are in phase with each other have no e?ect whatever. It is believed that a better understanding of the 10 action will be had from the following explana The reason that a low-pass ?lter may be used at. all is due to the fact that the characteristic frequencies. of the electroencephalogram signals lie sufficiently below 60 cycles for a ?lter to dis criminate between said signals and the hum without seriously distorting the wave shape of the signals, (60 cycles being the frequency of the power supply). Now it is evident that the two active electrodes of the subject or patient will pick up potentials, and that these potentials, if of bioelectric origin, will be different from each other in phase be cause the electrodes are situated over di?erent groups of cells. Interfering voltages, however, such as those proceeding from commercial fre-. quency electrostatic ?elds, will tend to affect the entire semi-conducting substance of the body simultaneously, and therefore the interfering I voltages as picked up by the electrodes will be in tion of the action of the unbalanced in-put sig nals. These signals pass through the ?rst stage of the ampli?er with substantially no change in - form other than their amplitude. For example, if a positive signal is delivered to one grid only of the ?rst tube in the ?rst stage, and no signal is‘ impressed upon the grid of the second tube in the. ?rst stage, an ampli?ed negative signal will result on the ?rst grid of the second stage, and no signal with reference to ground on the second grid of the. second stage. Now the negative sig nal on the second grid in stage two, reduces the cathode current for this tube. The cathodes, be ing tied together, will simultaneously go nega tive in voltage. Since however, grid. two of stage two remains at ground potential, this means that cathode two will go negative with respect to its former quiescent potential relative to grid two. There fore, as far‘ as tube two is concerned, grid two has received the equivalent of a positive signal; therefore, plate-current two will increase and in phase with each other. the output circuit the signal delivered to the‘ next In the ?rst. stage of the ampli?er, which is a stage will not only show a positive ampli?ed volt simple resistance-capacity coupled circuit, no age appearing on grid one of stage three, but discrimination or differentiation. between either I" also a negative and somewhat smaller voltage on type of potentials or signals referred to hereinbe grid two of stage three. The process repeats fore is attempted. In the second stage, however, until at the end of the cascade the plate-currents it will be noted that neither the cathode-resistor in the two tubes making up the fourth stage show nor the-screen supply resistor are by-passed. As long as the incoming signals from the first stage 40 very nearly equal and opposite changes in spite of the fact that the input signal has been applied are equal and opposite (symmetric and out-of to one side of the ampli?er. The screen circuits phase with each other) the cathode and screen of the pentode stages act to enhance this e?ect. currents in each tube will change by equal and It will be borne in mind that the selector opposite amounts and there will be no net change in the total cathode and screen currents of the pair. Assuming, however, that there are ‘also po switches when set for any two given electrodes, ' at the same time ground the other eight elec trodes and‘ the ‘latter are therefore inactive. These eight electrodes are obviously more or less tentials in phase with each other reaching both short-circuited together by the action of the grids, it will be obvious that this voltage will af-, fect both the screen and the cathode currents 50 switch, but the algebraic sum of their potentials is a. de?nite quantity with respect to the in the same direction, and thereby produce a algebraic sum of the potentials of the two active change in the cathode and screen voltage which electrodes. The action currents originating will tend to degenerate. or reduce the last men beneath these eight inactive electrodes are im tioned potentials. This continues on in the third and fourth stages with the result that while the 55 pressed all in- the same phase upon both input grids simultaneously. However, the ampli?er signals which are out of phase with each other degenerates these signals and they do not appear are ampli?ed more and more, the signals which at all in the oscillogram record. are in phase with each other are relatively am My new and improved electroencephalograph pli?ed less and less. The e?ect is further en device, referring now to Figure 1, is housed in a hanced by using resistors in the cathode circuits 60 metal case 23’ which has hinged thereto a cover as large as possible, consistent with keeping the 2!. A suitable lock 2?. is provided for securing operation of the tubes on a linear characteristic. the cover in its closed. position. The top panel Since no tubes except those in the last stage re 23. is secured in position in the metal case 28 by ceive signals large enough to operate them over of screws 24. Removably mounted in the an appreciable percentage of the total linear 65 means panel is a record carrying a feeding device gen plate characteristic, the cathode-resistor and consequently the grid~bias may be made much v erally designated by the numeral 25, which will presently’ be described in detail particularly in larger than would be possible were it necessary connection with Figures 3. '7, 8, and. 9. to obtain full plate swing in all of the stages. The controls and the recording head carried‘ It will be noted that in the circuits of my de 70 on the panel 23 will be described in connection vice I do not employ the obvious subterfuge of with the circuit diagram in Figure 10. using greater cathode-resistors and compensat The circuit is more or less conventionally shown ing for the excessive grid-bias produced by the in Figure 10, and the grids of the input tubes use of a negative voltage supply on its grounded end. Of course the use of such an expedient 75 29, 2-1, are extended and connected to switching 2,409,033 devices; the ‘grid of the tube 26, for example, is connected to the switch-arm 28 of the switch 29, and the grid of the tube 21 is connected to the switch-arm 30. Leads from the switch 29 are connected to a socket 32 mounted on the panel 23. This socket has contacts corresponding to all ten of the switch positions, and in addition has a ground connection so that when a plug, such as the plug 33, is ‘plugged into said socket, connections 8 bias cell 4| and a push-button key 42. With the switches in this position, the key 42 may be de pressed to record a standard, as will hereinafter be described. When the switches are turned to the “12” posi tion, which is marked “Cortex" on the panel 23, the grids of the tubes 26 and 2'! are connected to a jack 43, so that when a plug 44 is inserted in this jack, said plug being connected to a cortical are made to all ten of the leads in the cable 34— electrode 45, which will presently be described, the shielding 35 on the cable is at the same time the contacts of the cortical electrode are directly connected to said grids. connected to ground. This cable, as well as the one shown in Figure 13. will presently be de When the plug. is inserted in the jack 43, the scribed. 15 contact arms '16, 41, are moved away from each The switch 29 is in effect a two-gang switch, other and away from the contact arms 48, 49, respectively, which are connected together and and in one gang the contact arm 28 establishes contact with contact points to which the leads to to ground. In other words, the jack contacts 46 the socket 32 are connected. These leads are and 47 are normally connected together and to extended to the outside rim 29a of the second 20 ground until the plug is inserted. gang of the switch, and leads from the contact The cathodes of the the two triodes in the input points of the ?rst gang of the switch 3| are con stage, it will be noted, are connected directly to nected to the inside rim 29b of the second gang ground without any bias resistor. The ?lter of the first switch. chokes between the third and fourth stages re— The rotor 29 is ganged to the switch arm 28, so 25 ferred to are designated by the numerals 50 and 5!. that as the rotor 29c moves in unison with the arm 28, points on the switch 3| are connected to In the plate circuit of the ?rst and second corresponding points of the switch 29. Ganged stages of the ampli?er I have introduced a voltage to the switch 3| is a second gang wherein the regulator tube 52 for removingr the 69 or 120 cycle outer rim 3|a is connected to the contact points ripple from the ?lters following the recti?er and on the inner rim 29b of the ?rst switch, and for maintaining constant plate voltage on these two stages. wherein the contact points on the inner rim 1Hb are connected to ground. It will be noted that on the rotor 29° no bridg~ For controlling the sensitivity of the instru ment, I have introduced a double potentiometer ing contact appears in the position 36 (which in designated as R8. The movable arms thereof are the position shown represents a “7” position of connected to the grids of the tubes 53, 54, in the third stage. The shaft 61 of this double poten the switch 29). It will also be noted that the rotor 31 of the switch 3| has no bridging contact tiometer carries a knob 68 which operates over at the position 38 (which in the position shown a graduated scale 69 on the panel 23 by means of represents a “4” position of the switch 3|). 40 which the sensitivity of the device may be con With the switches in the positions shown, the trolled. grid of the tube 2'1 is connected to the conductor The resistance elements of these potentiome 4 in the cable, (to be presently described), and ters are each grounded at one end, and the other ends are capacitatively coupled to the plates of the grid of the tube 26 is connected to the con ductor '| in the said cable. It will also be noted that the inner rim 3|b has no ground connection at the vpoint 40; therefore, the “7" switch point on 3| is not grounded, and due to the absence of bridging connection at 38, the "4” position on the switch stage 3|a is also not grounded. However, all of the other switch positions on the switch 3| are grounded. Now. looking at switch 29, it will be seen that the wire connecting the “7” position on the switch 3| to the inner rim 2917 is not connected to the position “7” on switch 29 due to the ab sence of the bridging connection in the posi tion 36. The result is that electrodes connected to leads 1 and 4 (of the cable 34, Figure 1) are directly connected to the input grids of the tubes 26 and 21, and all of the other electrodes are grounded, and I may turn the switches 29 and 3| to other positions to place the grids under the in?uence of other electrodes at will, all other electrodes than the pair selected being automati cally grounded. Although only ten points have been described on the switches, they actually have twelve posi the tubes 55, 56, in the second stage by means of condensers C3, C4, respectively. The tubes 51, 58, in the fourth or last stage have their outputs connected to the moving coil 59 which drives the stylus 59, The ?eld struc ture 6| of the stylus drive is excited by the wind ing 62 which receives its current from the recti her 63, The stylus traverses the recording tape later ally on a rotating drum 64 which is driven by a synchronous motor 65 with suitable gear reduc tion therebetween. A switch 66 has three posi tions: (1) an “off” position; (2) a “ready” posi tion; and (3) a “recording” position. When the switch 66 is in the “ready” position, the heaters in all the tubes are supplied with current, but the exciting coil 52 is not energized, nor is the synchronous motor 65 energized. When the switch is in this position, the pilot light which is green, lights up. When the switch 66 is moved to the “record" position, the circuits are energized, including the exciting coil 62 and the synchronous motor 65; the pilot light PR, which is red, also lights up, and the pilot light PG is extinguished, tions, as will be seen in Figure 10. When both switch arms are positioned in the “11” position (which is marked “Standard” on the panel 23), the grids of the tubes 26 and 21 are connected to carrying a spring clip 1 2 is secured to the shield for connecting to a Water pipe or radiator. The the standardizing circuit 39. This circuit includes electrostatic shields 13, T4, are provided between The power cord 70 is shielded and a wire 1| 5 series of resistors R19 to R22 inclusive, a grid 75 the primaries and secondaries of the power trans formers. aaoaosa V The following is a list of ‘the values of the com ponents indicated on the circuit diagram: R1 ____ohms__ 100,000 R24 _____ohms__ 2,000 Re R3 R4 R5 ____do____ 100,000 _____megohm'__ 1/2 _______ __do____ 1A; _______ __do____ 1/2 . R25 R26 R27 R28 ____ “do-..” 6,000 ____ __do____ 200 ____megohm__ 1/2 ____ohms__ 25,000 Re _______ __do____ 1/2 C1 __microfarad__ 1 R7 _______ __do.____ 1/_> C2 ______ __d0____ 1 Rs, dual C3 ______ __do____ 1 potenti_ 10 9 .. - C4 upon which is mounted a roller 90 which may have a frictionless bearing between itself and said shaft. Spring means 9| urges the arm 89 in a counter-clockwise direction, as viewed in Figure 7. As may be seen in Figure 3, the free end of the tape extends from the roll 83 underneath and around the rotating drum 64 and across the ?at mid-portion of the panel. This may also be seen in Figure l. The roller 90 acts as a pressure roller 10 for holding the record in engagement with the knurled surface of the rotating drum 64. Referring now to Figure 9, the panel 23 has a hole 92 formed therein. This hole may also be seen in. Figure 7. A bracket 93 is secured to the ______ __do___a .02 15 panel 23 by means of screws 94 having nuts 95 ______ __d0____ .02 on the bottoms thereof. Mounted on the bracket ______ _._d0______ .02 93 is a vertical plate 96. Mounted on the vertical ______ __do____ .02 plate 96 is an L-shaped member 91 and a second ______ _..d0____ .1 L-shaped member 98; the legs of the Us of these _____ l_-_do__-_ 40 20 members face each other, thereby forming a slot '_ _____ __do_l__ l ometer ___do_ l/g-Vg R9 _______ __do____ 1/2 C5 ______ __~do____ C6 ________d0____ 1 1 R10 ______ __d0____ 1/z R11 _____ __ohms__ 50 R12 ______ __d0_____ 50 C7 C8» C9 R13 __-__megohm__ 1/2 C10 R14 ______ -_do____ 1/2 C11 R15 R16 R17 R18 C12 C13 ______ _fdo____ 60 C14 ______ __do____ 60 C15 ______ __d0____ 10 therebetween, the slot being designated by the C16 C11 ter and form a working ?t in the slot 99. ___ohms__ 15,000 ____do____ 750,000 _a__do____ 20,000 ____d0_..__ 50,000 R19 _____megohms__ 3 R20 ____ __ohms__ 150 ______ __do____ 15 ______ __do____ 10 numeral 99. The panel 23 is notched to conform to the slot '99 so that the vertical arm 18 may en V In threading up the tape, the free end 832 R21 _______do____ 150 C18 ______ .._d0____ 15 should project to the right from off the top of R22 _____megohms__ 3 C19 ______ __d0____ 5 the reel, as seen in Figure 7 ; ?rst holding the free R22 ____ohms__ 50,000 end of the tape on the reel to prevent it from unwinding, place the arm 18 (which may be L1—2220 henries; 2B and 21 Type 6F5; 52-VR 150-30 30 termed a slide) in the slot 99 between the guides L2-2220 henries; 55 and 56 Type 6SJ7; V1o~—5Z4 L3-—i5 henries; “53 and 54 Type 6SF5; V11--6X5 L1~l5 henries; 51 and 58 Type 6P6; 63-5T4. I do not wish to be bound by these values as they are merely given by way of example, and it must be distinctly understood that many changes made in the values of these components in the circuits may be made without departing from the 91, 98, and lower the panel partially. Then draw the free end of the tape underneath the rotating drum 64 and bring it around the drum and un derneath the stylus 60, which may be lifted to facilitate the operation. Then carry the free end of the tape to the left, as viewed in Figures 3 and 9, underneath the roller 90 and over the ?at sur face of the panel 16. Now lower the panel 16 until it comes in contact with the panel 23, taking spirit of the invention. ‘ 40 up the slack of the tape while doing so by drawing A fuse 15 is connected in series with one side it toward the left. When the tape is properly of the line feeding the primaries of the power threaded, it will have the appearance shown in transformers. This fuse is mounted in a cup shaped holder which projects downwardly from Figure 1. With the cable 34 connected to the machine by inserting the plug 33 in the socket 32, the ten cap which may be unscrewed for replacing blown conductors having metallic disc electrodes I00 fuses. connected thereto, are by this operation con In connection with the recording device 25, the nected to the switches 28 and 30, and the cable carrying the ground clip [01 is connected to panel 16, preferably formed of insulation, has secured thereto a bracket 17 (see Figures 7 and 50 ground by means of the side lug I02 on the plug 8) and a vertical arm 18 is secured to the bracket 33. The disc electrodes are numbered 1 to 10, in 11 in any suitable manner, for example by means clusive, corresponding to the switch positions 1 of the rivets T9. The vertical arm 18 has a por to 10, inclusive. tion 80 thereof extending above the surface of In attaching these disc electrodes to the scalp, the panel 70, and a hole 8! is formed therein to 55 the operation is facilitated if the subject is seated facilitate the removal of the panel and its con in a chair. For routine exploration, the follow-~ tents from the machine, as will hereinafter be ing scheme may be adopted: described. Electrode: A spindle 82 projects laterally from the arm # 1. Left frontal 18 and serves as a support for a roll 83 of record 60 I 2. Right frontal ~ tape. Suitable downwardly depending rods 84 3. Left parietal and 65 are mounted on the panel 16 and serve to 4. Right parietal align and guide the tape 83a toward the rotating 5. Left occipital drum 24. The panel 16 has a notch 85 formed 6. Right‘ occipital therein to accommodate the rotating drum 64 65 ‘ 9. Lobe of left ear which is provided for feeding the tape at a uni '1 0. Lobe of right ear form rate past the recording stylus 80. Electrodes #7 and #8 are placed in the vicinity The rotating drum 64, as hereinbefore pointed of any area from which other neurological signs out, is driven by a synchronous motor through . suitable gearing, and is of such diameter that 70 are suspect. I The electrodes are placed in contact with the the speed of the record tape past the stylus is scalp and collodion of a fairly heavy consistency exactly three centimeters per second. is applied to the skin and to the edges of the A boss 8‘! mounted on the panel 16 carries a electrodes with a medicine dropper. The drying stud 88 upon which an arm 89 is pivotally mount ed. On the end of the pivoted arm 89 is a shaft 75 of the collodion is greatly accelerated by means the panel 23 like a well, and is covered with a 11 2,409,033; of a current of warm air from an ordinary elec tric hair dryer. The electrodes may be held ?rmly in contact with the skin by means of a pencil point or an orange stick held in contact with the concave de pressions in the surfaces of the electrodes until the collodion sets. When the electrodes are se cured in place, the patient is transferred to an a 12 have been placed in the neighborhood of sus pected lesions or other areas of particular inter est. Then records are taken between various pairs of electrodes on the scalp, additional ones being applied if necessary, in order sharply to delineate any focus of unusual electrical gravity. For these readings both selector switches are of course used to connect the amplifier to any pair. examining table or cot-bed, and a low pillow may Never place arm 28 and arm 30 both on the same be placed under the patient’s head. The cot or electrode number. examining table may have a thin pad mattress Set the sensitivity control 68, to begin with, at, and the ground clip Hll should be connected to say, “70” on its scale 69. Wait an additional few the examining table or cot. If painted, the paint seconds for the ampli?er again to become sta should be scraped off so that the clean metal bilized. to metal contact is established between the 15 Now turn the starting switch 66 to “Record." ground clip and the metal of the examining table The red panel lamp PR will now light and the or the springs of the cot. It is thoroughly advisable, particularly for the tape draw-off motor will start feeding the tape "33a and the stylus 60 will begin to draw a black beginner, to make an electrical test of the resist line on the paper. If the tape begins to feed ance of the electrode contacts, because the ap 20 crookedly, return the switch 66 to “Ready,” pull paratus cannot function if there is a broken or the free end of the tape to straighten, then switch poor contact in the path of a current from the to “Record” again. The stylus 60 will move side brain to the ampli?er. The resistance of the elec wise, recording the electroencephalogram. The trode contacts is measured between any two elec stylus is electri?ed when the starting switch is trodes and should average between 3,000 ohms to on “Record” and will give a slight unpleasant a maximum of not more than 15,000 ohms. A shock if touched. higher resistance than this limit would have the twofold effect of reducing the sensitivity of the apparatus and of permitting extraneous inter ference to get into the ampli?er and obscure the record. If a higher resistance is noted between any If the tracing is not wide enough-that is, has not enough amplitude to give a clear recognizable and easily analyzed graph-very slowly increase the sensitivity by turning the knob 68 toward 100 on scale 69. If on the other hand, the stylus vibrates with such a great amplitude that it hits pair of electrodes, the defective member of the the internal stops at the ends of its limits of pair may be determined by checking each one travel, reduce slowly the sensitivity to obtain a separately with another electrode known to have 35 suitable tracing. Notations on the leads used, good contact. The electrode with the defective experimental conditions, or events of interest can contact is then taken off and carefully re-applied. be made directly in pencil on the tape. Before applying any electrodes to the scalp, a To record from other leads, ?rst stop the record material for reducing the surface contact resist ance may be rubbed into the scalp. All stray 40 by turning the starting switch back to “Ready." Then change the‘lead switches to the next pair wires and other apparatus should be removed of electrodes selected. Wait a few seconds for from the vicinity of the patient and no person stabilization of the ampli?er before turning back should touch. the patient or approach within to “Record.” three feet of the patient during a recording. At the conclusion of the test, remove the elec Make sure that the tape 83a can run freely over 45 trode tips from the jack and take the electrodes the motor drum 64 and that it comes off to the from the head of the subject by softening the left straight and smoothly across the Bakelite collodion with acetone. Clean the electrodes by panel 16; pull it lightly to be certain that all slack underneath the panel is taken up. washing with alcohol or ether to dissolve the Make sure that the stylus 60 is at rest in the 50 collodion and with water to remove any traces of center of the paper. If it stays at one side, push dried electrode paste. Store the electrodes care fully by fastening them again to the card on it over hard enough so that when it is released it will remain in the middle. which they are shipped or by hanging them from Turn the starting switch 66 to “Ready.” The a notched board attached to the wall so that the green panel light PG will show that the ampli- 55 wires will not become tangled. fler is turned on, and that the tubes are warming In different areas and in different subjects, up. Allow them three minutes to become stabi there is awide variation in the amplitude of the lized before attempting to record. waves found. It is, therefore, necessary, as (ii For a routine examination by a method now rected above, to adjust the sensitivity of the re coming to be accepted as standard the procedure 60 cording mechanism to obtain graphs which, on is to make records of the left frontal, parietal, the one hand, must be large enough to analyze and occipital areas with respect to an indifferent and which, on the other hand, must be small electrode, and the same for the right side. To enough to be recorded within the limits of Width obtain these records with the selector switch, set of the tape. It is desirable, however, to know at arm 30 on 9 while arm 28 is placed successively 65 all times the absolute magnitude of the waves on electrodes 1, 3, and 5, then set arm 30 on 10, 7 being recorded. while arm 28 is placed on electrodes 2, 4, and 6. By this means the indifferent electrode, which is in contact with the lobe of the ear, is on the same To ?nd this, for any setting of the sensitivity control 68, turn the starting switch 66 to “Ready,” then set arm 28 and arm 30 both on “Standard." side of the head as the active electrodes, with 70 Wait a few seconds, then turn the starting switch the result that some cross potentials from deeper 66 back to “Record.” Now depress and release structures are eliminated in the recording. the black button 42 marked “Standardize.” The The procedure outlined gives records of what stylus 60 will make a vertical mark upwards when are known as resting potentials. Further records should be taken with electrodes 7 and 8, which 75 the button is depressed and back again when the button is released. The length of this verti 2,409,0ser l3? calf line upwards is the response of the stylus‘ to a 50 microvolt signal. It is also essential to note the speed of the tape in order that the frequency of the waves may be computed. In this’ instrument, the speed is ?xed at exactly 3 centimeters per second. Time rela tions may instantly be found by measuring the I4? plug has a hole I3I formed ‘therein which forms a socket to receive the head of the electrode, and spring members I32 are provided for reducing the contact resistance between the plug and the head. The headset may be applied to the subject, and the curved members III, II2, adjusted to desir able positions and secured by means of the thumb tape with a common. metric ruler. nuts H3, and the electrodes may be applied to Special electrodes are available for taking elec trocorticograms from the exposed cortex at op 10' any desired combinations of tap holes Hi]: after which the plugs it") may be connected to the eration under'sterile conditions, for taking elec electrodes and the socket I23. plugged in. The tromyograms from single muscular units, for tak procedure may follow‘ the procedure. described ing gross electromyograms from the surface of above in connection with the. cable 34‘. the skin over the skeletal. muscles, etc. To use The recording mechanism includes four’ pole 15 these-electrodes, set both arm 28 and arm 30 at pieces and the powerful direct current electro positions marked “Cortex” and plug the special magnet 62 energizes these pole pieces. The mov electrode plug ?lllvinto the jack marked 43, “Corti ing armature coil is wound in two sections, and cal electrodes.” both are placed over the armature so that ity is The cortical electrode 65 includes a- holder Hi3, magnetized virtually by the algebraic sum of the the outer end I-M of which has a socket formed two currents of the two coils; one coil is placed therein so that the electrode I65 per so may be in the: plate circuit of the output tube 51, and the removed or inserted at will, thereby permitting other is placed in the circuit of the output tube the electrode to be properly removed and 58. The direct current plate components ?ow sterilized before being placed in contact with the through these coils in such a direction that the tissue. The end of the electrode includes a magnetic fields cancel. The coils are wound to grounded or “shield” area I86 surrounding the the correct load impedance for the output tubes. active or central core IN. The entire electrode The armature is returned to its central posi cord may also be sterilized. tlon by a pair of very strong phosphor-bronze In Figures 11 to 14, inclusive, another electrode helical tensionv springs, with the result that the“ and cable arrangement is shown. A‘ head-set, recording mechanism has exceedingly high e?i generally designated by the numeralv H38, in ciency and develops relatively enormous power. cludes a head band £59 which may be somewhat oval in form and larger than the average head. A plurality of tapped holes Ilil is vtermed in the band we enabling the operator to position the scalp electrodes, which will presently be described, in a verylarge number of combinations of po sitions. Pivotally secured to the band» its are curved members III, I.I2, which also contain a plurality of tapped holes HQ; The means for pivotally securing the members Ill, H2, to the band I88 are thumb screws [It whichenable the operator to. set the pair of members lII, H2, in any desired position with respect to each other and. with respect to the band I99. The electrodes, one of which is shown enlarged in Figure 14, consist of of a bushing IIrl having a hole therethrough which forms a working ?t with a shank H5. The bushing also includes an "The condenser Cls'an'd resistor R28 supply crit ical damping to they moving element. of the re cording mechanism. * The armature shaft‘ is. vertical, and at the top of. this shaft is a steel cross-head (not shown) which carries the stylus 60 by a right-angle ex tension of the latter which passes through the , head horizontally. The stylus is therefore moved through a horizontal are by the vibrations of the armature, but is held in contact with the record 83a by a helical torsion spring which urges it ver tically downwardly. The stylus is maintained at a positive poten-~ tial byuthe transformer ‘M and its associated recti?er and ?lter system, the current passing through the limiting resistor R23 to the stylus. The passage of the current produces an imme-~ outer threaded surface I I6 which ?ts the threaded 50 diately visible and permanent record on the pre pared electro-chemically sensitive tape 83*‘. This holes Ilii. The shank H5 carries a head which current‘ is turned on simultaneously with the cur has a groove H8 formed therein and the end H9 rent to the motor 65' and. the current to the field is rounded to permit the easy application of a plug thereto, which plug will. presently be de scribed. 82 by moving switch 66 to position “Record.” Condenser C11 s'erves‘to eliminate radio interfer ence caused by arcing to the stylus point. This’ On. the other end of the shank H5 is an elec recording system, without electri?cation of the trode I20, the end IZI of which is concave to stylus, is suitable for use with ‘a. prepared. waxm contain a material for reducing the contact re covered tape' sistance between’ the electrode and the scalp. A A shield I33 having an angular portion I35 is0-> spring.’ I22 extends from the left end of the bush 60 lates the ?rst three ampli?er stages from the ing, as viewed in Figure 14, to the electrode I20 fourth stage, the recti?ers and the. recording and urges the electrode to the left (which would devices. The shield I34 joins the shield I33 and be toward the scalp when the headset is in po is positioned between the recti?ers and the re sition on the patient) ., Referring now to Figure 13, the plug I231 is like 65 cording device. I‘. have also made electroencephalographs for the plug 33 previously described, and includes ten multi-channel work which employed a, plurality contact pins I124 which fit the socket 32, and a of ampli?ers (as many ampli?ers as channels de ground lug I25 establishes a ground contact with sired). Figure 15 shows an arrangement for a. thev chassis. The cable I26 is covered with a shielding I21, and‘ the shielding and the ground elip I23 are connected to the ground lug I25. The ends of the‘ wires of the cable carry tabs 129 which’ bear numbers corresponding to the posi tionsofv the’ switches 28, and 30. Each conductor. has-secured to the end thereof a plug. I3.EI.- This 75. multi-channel electroencephalograph, portions of the ampli?er being omitted to avoid duplication. In this arrangement the multi-point switches employed have seventeen points and siXteen- ac tive positions, the seventeenth point being con nected to ground. < ~ 2,409,033 15 16 The switches in channel I are designated by mean potentials which are out-oI-phase with the numerals I36 and I31, and the switches in each other or which are of opposite phase, as the channel II are designated by the numerals I38 device deals with potentials which are primarily and I39. The midpoint of the switch I36 is con picked up by not less than two active electrodes on nected to the grid of the input tube 261’, and the a biological preparation. One end of the group midpoint of the switch I31 is connected to the of cells in the biological preparation may be grid of the input tube 21‘). The input tubes 26b negative with respect to the other end of the and 21'’ are exactly like the input tubes 26 and same group when a bio-electric potential is being 21 shown in Figure 10, and the balance of the circuit associated with the tubes 26” and 21b is 10 picked up by the electrodes. The term “potentials in phase with each other," exactly like the circuits associated with the tubes “potentials in phase with one another," or “in 26 and 21 in Figure 10, and therefore they need terfering potentials” used herein is taken to not be repeated in Figure 15 or further described. mean any extraneous or interfering potentials, In channel II the midpoint of the switch I38 such as static, both man made and natural. It is connected to the grid of input tube 26c and the will be appreciated that the bio-electric potentials midpoint of the switch I39 is connected to the are very small and that all interfering potentials grid of the input tube 27°, the rest of the ampli are very great in comparison thereto. ?er also being like that shown in Figure 10. It will be evident from a study of the above The stylus controlled by the channel I and the speci?cation that I construct my ampli?er so stylus controlled by the channel II are placed that each stage therein tends to cancel out inter side by side in reasonably close spaced relation fering potenials and at the same time to amplify to each other. The tape used with the two sty bio-electric potentials (potentials which are out luses is substantially wider than the tape 83“; of phase with one another or with each other) also the rotating drum is likewise Wider than the and that as a means of catching and shoving out rotating drum 64; but the driving arrangement is any interfering potentials which have not been identical to that shown at 65 in Figures 3 and 4. previously balanced out or which are instituted In Figure 15 a further modi?cation is shown, in the ampli?er itself, I provide ?lter chokes be which modi?cation may likewise be applied to tween the output of the next to the last stage the single channel electroencephalograph herein and the ?nal stage. before described. Instead of short-circuiting and Although I have herein shown and described grounding the switch contacts not in use, and by way of illustration 9, device for producing elec thereby grounding the electrodes to which they troencephalograms which obviate the necessity are connected, this shorting feature may be for shielding the instrument and/or the patient by omitted and the switch contacts may all be con means of a “Faraday cage” or the like, said device nected to ground through resistors. In multi-channel devices, I have successfully employed resistors for this purpose each having also being capable of making electrocardiograms, I do not wish to be bound by the exact arrange ments and the speci?c values of elements herein a resistance of 100,000 ohms. This results in a shown and described, as it is obvious that many large number of parallel grounded resistors, and the ?nal effect in interference suppression is 40 changes may be made in the apparatus shown without departing from the spirit of the invention found to be substantially the equivalent of the as set forth in the annexed claims. shorting arrangement shown in Figure 10. What is claimed is: Another advantage is that with this arrange 1. In a device for making recordings of po ment the switches are simpler and therefore less expensive and less liable to get out of order. tentials generated in the human brain, a multi In Figure 15, I have shown the leads running Jr. stage push-pull ampli?er, a plurality of contact members, means to hold said contact members from the switches as terminating in individual in predetermined or de?nite positions in con binding posts, but it is obvious that these could tact with the scalp of a patient, switching be connected to sockets in the manner shown in Figure 10. means, conductors for connecting said contact 1:: members to said switching means, connections The device herein described may also be used between said switching means and the control as an electrocardiograph, in fact it is believed grids of the ?rst stage of said ampli?er, said to be the only device ever built which will suc switching means being adapted to selectively con cessfully produce an electrocardiogram not re nect said grids to any two of said conductors quiring photographic development for the pres ‘ and to ground all of the other conductors, ervation of the record. Notwithstanding anal shielding means surrounding said conductors and yses which indicate the desirability of a fre connected to ground potential, a metallic table or quency range up to at least 240 cycles per sec the like for supporting said patient, a thin in ond, my device turns out, even with the ?lters sulating pad between said table and said patient, described in the circuit, very creditable records. means for connecting said support to ground By cutting out the ?lters, I obtain a higher fre potential :thereby reducing the impedance of quency range which admittedly is advantageous. In using the apparatus as an electrocardio graph, it is merely necessary to position the elec trodes to pick up heart potentials instead of brain potentials. When the device is used as an elec trocardiograph, the grounding of the electrodes is not necessary in view of the fact that the amplitude of the electrocardiogram is so much greater than that of the electroencephalogram that the precautions of hum ?ltering and ground ing of additional electrodes are entirely unneces sary. ‘ said tissue and leading off interfering potentials to ground, and a recording device connected to the output of said ampli?er for recording poten ' tials picked up by said contact members from said patient after the same have been ampli?ed. 2. In an electroencephalograph, a panel, an ampli?er beneath said panel, a recording device including a motor-driven drum, a recording tape embracing said drum in driven relation thereto, a stylus carried on said panel in contact with the tape on said drum, said stylus being motivated by the output current of said ampli?er; and a, In this speci?cation and in the appended claims, the term “bio-electric potentials” are taken to 75 control switch on said panel having a first posi tion wherein all circuits are “dead,” a second po 2,409,033 ‘ 17 sition wherein the heaters of the tubes in said ampli?er are supplied with energy, and a third position wherein said ampli?er is rendered fully operable and ‘the motor for driving said drum is supplied with energy for moving said tape past said stylus. 3. In an electroencephalcgraph, in combination, a head-set including a substantially rigid oval 18 to- ground and other non by-passed resistors in series with the screen grids of said several stages for causing bio-electric or out-of-phase poten tials to :be ampli?ed and for effecting the nulli ?cation or balancing out of in-phase‘ or interfer ence potentials, means for connecting said ampli ?er to a source of alternating current, ?lter choke means connected between the output of one of said stages and the input of the next suc band having a plurality of tapped holes formed therein in spaced relation to each other, at least 10 ceeding stage for deleting interference poten tials which the previous stages have failed to one arcuate member being pivotally supported on balance out, together with interfering potentials said band and also carrying a plurality of spaced originating in the ampli?er itself and the alter tapped holes, a plurality of electrodes adapted to nating current hum, recording means connected be selectively positioned in any of said tapped holes and each being comprised of a scalp-engag 15 to the output of said ampli?er, a panel, a pair Of multi-c'ontact switches on said panel con ing member, a shank, and a terminal end, said nected to the input tubes of said ampli?er, a shank carrying a spring and a threaded bushing plurality of shielded conductors connected to the having threads to match said tapped holes, said contacts of said switches, means for connecting spring being so positioned on said shank as to urge its scalp-engaging member into resilient 20 said cables to sources of bio-electric or out-of phase potentials, and connections between contact with ‘the scalp, and a cable connected to ground potential and said switches, said last said electroencephalograph and carrying sockets connections being arranged to ground all of the engaging said terminal ends. contacts on said switches except one on each 4. In an electroencephalograph, in combina tion, a head-set including a substantially rigid 25 switch, whereby a single selected contact on each switch may be ungrounded and connected to one oval band having a plurality of tapped holes of the input tubes of said ampli?er. formed therein in spaced relation to each other, rI. In an electroencephalograph, a multistage a plurality of arcuate members pivotally con push-pull thermionic ampli?er including resis nected to said band and each also carrying a plu rality of spaced tapped holes, whereby said arcu 30 tors and capacitors interconnecting the tubes of said stages and non by-passed resistors connect ate members may be given any desired positions ing the cathodes in several of said stages directly with respect to each other and to said band, to ground and other non by-passed resistors in means to clamp said arcuate members in any de series‘ with the screen grids of said several stages sired positions; a plurality of electrodes adapted for'causing bio-electric or out-of-phase poten to be selectively positioned in any of said tapped tials to be ampli?ed and for effecting the nulli holes, and each being comprised of a scalp-en ?cation or balancing out of in-phase or inter gaging member, a shank, and a terminal end, said ference potentials, means for connecting said shank carrying a spring and a threaded bushing ampli?er to a source of alternating current, ?lter having threads to match said tapped holes, said spring being so positioned on said shank as to 40 choke means connected between the output of one of said stages and the input of the next suc urge its scalp-engaging member into resilient ceeding stage for deleting interference potentials contact with the scalp; and a cable connected to which the previous stages have failed to balance said electroencephalagraph and carrying sockets engaging said terminal ends. out, together with interfering potentials origi 5. In an electroencephalograph, a multistage 45 nating in the ampli?er itself and the alternating current hum, recording means connected to the push-pull thermionic ampli?er including resist output of said ampli?er, a cortical electrode in ors and capacitors interconnecting said stages cluding a central electrode surrounded by a and non-by-passed resistors included in circuits with the cathodes and screen grids of several 50 grounded, shielded area, both being positioned in, contact with brain tissue, said central elec of said stages themselves for causing bio-electric trode being adapted to be influenced by brain as out-of-phase potentials to be ampli?ed and potentials, conductors connected to said cortical for effecting the nulli?cation or balancing out of electrode and to the input tubes of said ampli in-phase or interference potentials, means for connecting said ampli?er to a source of alter 55 ?er, and a shield about said conductors and con nected to ground potential. nating current, ?lter choke means connected be 8. In an electroencephalograph, a multistage tween the output of one of said stages and the push-pull thermionic ampli?er including resis input of the next succeeding stage for deleting in tors and capacitors interconnecting the tubes in terference potentials which the previous stages have failed to balance out, together with inter 60 said stages and non by-passed resistors connect ing the cathodes in several of said stages directly‘ fering potentials originating in the ampli?er it to ground and other non by-passed resistors in self and the alternating current hum, a pair of series with the screen grid circuits of said several multi-contact switches connected to the input stages for causing bio-electric or out-of-phase tubes of said ampli?er, a plurality of conductors potentials to be ampli?ed and for effecting the connected to said switches, whereby the inputs 65 nulli?cation or balancing out of in-phase or in of said tubes may be selectively connected to any terference potentials, said ampli?er also includ two of said conductors and bi0~electric or out-of ing resistors between the input grids of the ?rst phase potentials carried by any combination of stage of said ampli?er and ground potential for two of said conductors selected by said switches limiting the impedance across which interfering may be delivered to the input of said ampli?er. 70 6. In an electroencephalograph, a multistage push-pull thermionic ampli?er including resis ?elds may build up voltages, at ?lter choke be tween the input of the last stage and the output of the next preceding stage for deleting inter tors and capacitors interconnecting the tubes of ference potentials which the preceding stages said stages and non by-passed resistors connect have failed to balance out, together with inter 75 ing the cathodes in several of said stages directly 19 2,409,033 fering potentials originating in the ampli?er it self, recording means connected to the output of said ampli?er, a shielded cable having a plurality of conductors positioned on a biological prepara tion, selective switching means connected to said 20 ternating current hum, and recording means con nected to the output of said fourth stage. 11. In an electroencephalograph, a multi-stage push-pull ampli?er, a plurality of contact mem bers, means to hold said contact members in pre determined or definite positions in contact with ductors whereby different sources of bio-electric animal tissue, switching means, conductors for or out-of-phase potentials may be delivered to connecting said contact members to said switch said grids, and shortcircuiting means included in ing means, connection between said switching said switching means for grounding all said con 10 means and the control grids of the ?rst stage of tacts except the ones connected to said grids. said ampli?er, said switching means being 9. In an electroencephalograph, a four stage adapted to selectively connect said grids to any push-pull ampli?er including a power pack for of said conductors, shielding means surrounding amplifying potentials which are out of phase with said conductors and connected to ground poten each other and for nullifying potentials which 15 tial, the two tubes in said ?rst stage including are in phase with each other and termed inter cathodes which are directly connected to ground fering potentials, means for connecting said am potential thereby placing them intermediate the pli?er to a source of alternating current, con potentials on said control grids, a metallic table or nections between the control grids of the ?rst of the like for supporting said animal tissue, an in said stages and a source of potentials which are 20 sulating pad between said table and said animal out of phase with each other, means for shielding tissue, means for connecting said support to grids and having contacts connected to said con said connections from interfering ?elds, resist ground potential thereby reducing the impedance ance bridges from said grids to ground potential for limiting the impedance across which interfer ing ?elds may build up voltages, means for main taining constant plate voltages on said ?rst and the second of said stages, thereby preventing plate voltage changes which would be ampli?ed and appear as interfering signals in the third and fourth stages, at least one of said stages includ ing two anodes, two cathodes, two control grids and two screen grids, said control grids being capacitively coupled to the anodes of said ?rst stage, said cathodes and said screen grids employ ing non by-passed resistors, ?lter chokes between ; said third and said fourth stages for ?ltering out alternating current hum, and recording means connected to the output of said fourth stage. 10. In an electroencephalograph, a four stage push-Dull ampli?er for amplifying potentials which are out of phase with each other and termed bio-electric and for nullifying potentials which are in phase with each other and termed interfering potentials, a power pack for said am pli?er, means for connecting the same to an al ternating current power line, shielded conductors connected to the control grids of the input stage of said ampli?er and to a source of bio-electric potentials, terminals on said cables and included in the connections to said source, means for re ducing the contact resistance between said ter minals and said source, resistance bridges from said control grids to ground potential for de?ning the maximum impedance across which interfering ?elds may build up voltages, means for maintain ing constant plate voltages on said ?rst and the second of said stages, thereby preventing plate voltage changes which would be ampli?ed and of said tissue and leading off interfering poten tials to ground, said ampli?er including at least other stages each employing two anodes, two cathodes, two control grids and two screen grids, said control grids being capacitively coupled to the anodes of the preceeding stage, said cathodes and said screen grids employing non by-passed resistors for minimizing any tendency for changes to occur in the values of the net total cathode and screen currents when potentials which are out of phase with each other and termed bio-electric potentials are impressed on the grids of said ?rst stage, the cathode and screen currents in each stage both being adapted to be affected in the same direction when potentials which are in phase with each other and termed interfering potentials are impressed across the grids of said 40 ?rst stage whereby said ?rst mentioned potentials are ampli?ed and said last mentioned potentials are degenerated or reduced, and a recording device connected to the output of said ampli?er for recording potentials picked up by said contact members from said tissue after the same have been ampli?ed. ,. 12. In an electroencephalograph, a multistage push-pull thermionic tube ampli?er including re sistors and capacitors interconnecting said stages, at least two of said stages including screen grid tubes having their cathodes connected to ground potential through non by-passed resistors and having their screen grids connected through non by-passed resistors to the high potential source in said ampli?er for causing potentials respec tively opposite to each other in phase impressed on the input of said ampli?er to be ampli?ed, and for effecting the nulli?cation or balancing out of appear as interfering signals in the third and fourth stages, others of said stages each includ 60 potentials which are in phase with one another, said ampli?er also including ?lter choke means ing two anodes, two cathodes, two control grids between the output of one of said stages and the and two screen grids, said control grids being ca input of the next succeeding stage for deleting pacitively coupled to the anodes of said ?rst stage, those of said last mentioned potentials which the said cathodes and said screen grids employing preceding stages failed to balance out, together non by-passed resistors for diminishing the tend with other interfering potentials also in phase ency for changes to occur in the net total cathode with one another and originating in the ampli?er and screen currents when said bio-electric poten itself, recording means connected to the output tials are impressed on the control grids of said of said ampli?er, and shielded conductors for ?rst stage, the said cathode and screen currents leading said potentials that are respectively op both being affected in the same direction by said posite to each other in phase to the input of said interfering potentials impressed across the grids of the input stage, ?lter chokes between said third and said fourth stages for ?ltering out al ampli?er, LOVE'I'I‘ GARCEAU.