EXCHANGE THE EFFECT OF TEMPERATURE -ON THE CHANGE OF RESISTANCE OF BISMUTH FILMS IN A MAGNETIC FIELD HY LEON FRANCIS CURTISS A THESIS PRESENTED TO THE FACULTY OF THE GRADUATE SCHOOL OF CORNELL UNIVERSITY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY lA Reprinted from PHYSICAL REVIEW, pp. 255-274, Vol. XVIII, No. 4, October, 1921 Reprinted from the PHYSICAL RKVIRW. N.S., Vol. XVIII, No. 4, October, 1921, PHYSICAL PROPERTIES OF THIN METALLIC FILMS. 1 III. THE EFFECT OF TEMPERATURE ON THE CHANGE OF RESISTANCE OF BISMUTH FILMS IN A MAGNETIC FIELD. BY L. F. CURTISS. SYNOPSIS. Ageing of sputtered bismuth films. The films sputtered on glass between sputtered gold terminals, were sealed in glass tubes and kept in vacuo at 210-220 C. for a number of hours. As a result each film showed a gradual decrease of resistance amounting in all to about 40 per cent. Eventually the films became so thoroughly aged that temperature-resistance measurements in vacuo could be accurately reproduced over long periods. Increase of resistance of bismuth films in a magnetic field for temperatures from 190 to 230 C. The isothermal curves obtained for each film by varying the field up to 18,000 gauss, have the same form for all temperatures, differing only by constant factors, and closely resemble the corresponding curves for bismuth in bulk. The percentage increase of resistance is not proportional to the square of the field. It is greater the lower the temperature and the lower the resistance, that is, the thicker the film, reaching a value about 1/5 that of bismuth for the thickest film tested. Variation of resistance of bismuth films with temperature, 190 to 230 C. The curves are parabolic in form, showing a minimum around 150-200 C.; the lower the resistance of the film, the lower the temperature of minimum resistance. Electrically heated oil bath. // is suggested that bakelite is a suitable material for protecting the nichrome heater. THIS work continues the investigation of the behavior of the re- sistance of bismuth films in a magnetic field described in the second 2 paper of this series. A study of the change of resistance of these films in a magnetic field at various constant temperatures from near the melting point of bismuth down to liquid air temperatures has here been undertaken. 1 Third of a series of articles on this subject from the Physical Laboratory of Cornell University. This series of articles was made possible by a grant to Prof. F. K. Richtmyer from the Rumford Fund. 2 F. K. Richtmyer and L. F. Curtiss, PHYS. REV., 15, p. 465, 1920. 255 ''*: L - F - CURTISS - For this purpose thoroughly aged films were desirable, since only with films which have been subjected to a thorough heat treatment can reproducible results be expected, as is evident from the results of the previous work referred to above. Since the oil bath had not been entirely satisfactory because of its destructive action on the films, it was thought advisable to try some other means of protecting the films from the action of the air during the heating. The procedure which suggested itself was to place the films in flat glass tubes and to evacuate the tubes to as high a degree as possible with a Langmuir mercury vapor pump, at the same time heating them and the enclosed films to near the melting point of bismuth. This heating served two purposes. It gave the films a preliminary heat treatment and it also helped to free the film and tube from adsorbed gases, thus affording a greater protection to the film during subsequent heatings. ' The films were prepared in much the same way as for the previous work (loc. cit.}. The glass slides on which the metal was deposited were made somewhat narrower for convenience in inserting them into the flat glass tubes. The actual dimensions of the bismuth film itself, how- ever, were kept the same as those used formerly, i.e., 10 by 3 mm. The rotating sector was dispensed with and a kenotron 1 for rectifying the secondary current of the transformer was placed in series with the dis- charge jar. This improvement cut out the reverse current and eliminated its heating effect, thus rendering the rotating sector unnecessary for the preparation of films for the present purpose. It also made possible the sputtering of much thicker films than had previously been possible. For, as stated in the preceding paper, after sputtering a certain length of time the surface of the bismuth film becomes gray and powdery and further sputtering has little effect upon the resistance of the film, since the particles of metal which come down seem no longer to adhere to the metal already deposited firmly enough to become active in conducting the current. It was found that, this condition did not come about as soon when the kenotron was used. The successive steps in the manufacture of a film are illustrated in the diagrams in Fig. I, which are drawn to scale. First the glass slide was thoroughly cleaned chemically, rinsed with distilled water and dried. It was then covered across the middle squarely with a glass slide exactly i cm. wide with edges ground true and gold terminals were deposited on the exposed ends by cathodic sputtering, after which it had the appear- ance indicated in Fig. I, a. Then a shield of mica with a slot 3 mm. wide and long enough to overlap the gold terminals was placed on this blank and a bismuth film deposited, likewise by cathodic sputtering, No L 4? VIH 'J PHYSICAL PROPERTIES OF THIN METALLIC FILMS. 2^"/- along the center of the slide connecting the two gold terminals, as shown in Fig. i, &. A piece of fine bare copper wire was then wrapped tightly around each end as shown in Fig. I, c. A thin layer of hot paraffin was spread across the face of the film, near the ends of the bismuth, which prevented the copper sulphate solution, used in plating the wires in position, from creeping up on to the bismuth. This was later dissolved off with xylol. First one end, then the other, of the film was dipped into the plating bath and a thick coating of copper deposited on the gold film and copper wire, the projecting wires serving as leads for the current in each case so that it did not pass through the bismuth film. This plated terminal insured positive contact with the film at all temperatures and proved to be far superior to the clamps which had been used before for this purpose. The film at this stage had the appearance shown in Fig. i, d. Short pieces of Cunife wire were then welded onto the pro- jecting ends of the copper wires for sealing through the glass and then longer pieces of copper wire were welded onto the Cunife to serve as external leads after the films had been sealed into the glass tubes. The tubes were prepared from flat glass tubing by sealing a short piece of glass tubing at right angles near one end for attaching to the pumps. The film with its lead wires was then sealed in and the tube sealed on to the pumps and evacuated for several hours at about 210. C. with a Langmuir pump, after which it was sealed off. The general arrange- ment of the film within its evacuated tube is also shown in Fig. i. The Fig. 1. film was now further aged by heating for four hour periods to about 220 in an oil bath. These heatings were repeated four or five times for each film after which it had become relatively stable and further heating had very little effect upon its resistance at room temperature. 1 Kindly loaned by Prof. J. S. Shearer. T r " rSKCOND [SERIES. It is interesting to note that all films, without exception, decreased in resistance as a result of this heating. This is quite contrary to the effect produced by heating bismuth films in air, or even in an oil bath, since only in the case of two films out of a couple of dozen did such an effect occur in the previous work (loc. cit.}. Thus Film V 5 had a resistance of 36.7 ohms at room temperature after it had been sealed into a glass tube and the tube exhausted. After it had been subse- quently heated for four hours at about 220 its resistance at room tem- perature was only 21.7 ohms, thus decreasing about 41 per cent, in resistance during this treatment. Similarly Film V 13 had a resistance after being sealed into a tube of 30.8 ohms and after heating eight hours to about the same temperature it had a resistance of 17.8 ohms, or had decreased about 42 per cent, in resistance. Other films showed com- paratively smaller changes during this aging process and in consequence were less stable, and showed further decreases in resistance during sub- sequent measurements at the higher temperatures. Thus the resistance of Film V~9 was 10.9 ohms after having been sealed into a tube and after heating for four hours its resistance was 7.4 ohms, decreasing about 33 per cent.; similarly Film V-I4 had a resistance of 100.0 ohms after it was sealed into a glass tube and after heating eight hours its resistance was 83.6 ohms, decreasing only sixteen per cent. Before the final meas- urements on this film had been taken, however, it had decreased to about 55 ohms. For measurements in the field at room temperature and above, a brass container for an oil bath was constructed. It consisted of two brass tubes, each cut away on one side and joined by a narrower neck so as to fit in between the pole-pieces of the magnet. A step-bearing was fastened in the center of the bottom of each tube in which a vertical shaft carrying a propeller for stirring was mounted. These shafts were provided with pulleys at the top which were connected by means of idlers with an end- less belt which also passed around the pulley of a small motor. The shafts were so driven that one propeller forced the oil down one cylinder and the other pulled it up in the other cylinder so that a continuous circulation of oil was produced. Fig. 2 shows a plan of the container and illustrates its position with reference to the pole-pieces of the magnet. The bath was about 12 cm. deep and 22 cm. from the outside of one cylinder to the outside of the opposite cylinder. The distance between the pole-pieces was one centimeter. The bath was heated by a resistance element made by winding nichrome ribbon on a sheet of mica. It was found that some protection for the ribbon was necessary since the oil, although it was a good grade of gas-engine cylinder oil, carbonized badly PHYSICAL PROPERTIES OF THIN METALLIC FILMS. 259 at the higher temperatures and adhered to the ribbon, thus short- circuiting adjoining turns. Bakelite was found very satisfactory for this purpose. The current through the heating element could be con- trolled by rheostats so that the temperature could be brought to any desired value and held constant for considerable intervals of time. The temperatures were measured with a copper-advance thermo-couple made and calibrated in the usual way. Its e.m.f. was measured by a Leeds and Northrup potentiometer. Fig. 2. Since simultaneous measurements of field strength and resistance of the film were impracticable, a calibration of the magnet was made before the temperature bath was inserted between the pole-pieces. This was done by means of a bismuth spiral, calibrated for the purpose by the Bureau of Standards. Especial precautions were taken to keep the spiral during these measurements at the constant temperature at which it had been calibrated. By this means a curve between the current through the magnet coils and the strength of the resulting field in kilogauss was obtained which was used in determining the field throughout the experi- ments at room temperature and above. Control experiments showed that this method would give results which could be depended upon to within less than one per cent., which was deemed adequate for the present work. The current for the magnet was supplied by a set of twenty-six 15-ampere storage cells connected in series. The maximum current 'used in this part of the work was about seven amperes. A large Weston Laboratory Standard millivoltmeter and shunt were used in measuring the magnet current both during the calibration and throughout the subsequent experiments. For convenience, all the measuring instruments and controlling rheo- stats were mounted on a table so that they were all under the control of one operator. Thus all factors could be varied by a single person without shift of position during an experiment. This also made possible a simultaneous determination of the various quantities, eliminating 260 L. F. CURTISS. [SECOND [.SERIES. errors due to variation in the temperature or in the current through the magnet. In making the measurements the procedure was as follows: The film was placed in its position in the oil bath and the rheostats controlling the current through the heating element were adjusted. About twenty minutes were necessary for the temperature to become steady. As soon as the potentiometer readings indicated that this state had been reached, measurements of the resistance of the film were made at various values of the current through the magnet coils, checking the zero-field resistance of the film and the potentiometer reading after each measure- ment. It was found possible in all cases to hold the temperature steady enough so that the zero-field resistance of the film did not vary more than o.i per cent, during a series of measurements at a given temperature. The results of the measurements can best be presented by a discussion of the curves plotted from the data taken with typical films. A set of data for one of the films, V-5, is given below to illustrate the magni- tude of the quantities involved. FILM V-5. Potentiometer Reading, o.ooioio; Temperature, 23.5; Resistance of leads, 0.370 ohms. Magnet Current. Bridge Reading. Resistance of Film. dr drlr H (Gauss). 1.0 21.603 21.509 21.233 21.139 0.094 0.0044 2300 2.0 21.862 21.516 21.492 21.146 0.346 0.0164 4725 3.0 22.219 21.502 21.849 21.132 0.717 0.0339 7250 4.0 22.682 21.498 22.312 21.128 1.184 0.0561 9700 5.0 23.170 21.492 22.800 21.122 1.678 0.0795 11950 6.0 23.660 21.492 23.290 21.122 2.168 0.1028 14125 6.84 24.014 21.493 23.644 21.123 2.521 0.1196 15710 Another set of data for the same film at 230, the maximum tempera- ture at which measurements were made on this film, is given below also. PHYSICAL PROPERTIES OF THIN METALLIC FILMS. 26 1 Temperature variations are more likely to occur and the zero-field re- sistance of the film is not as constant as in the set taken at room tempera- ture. However the variations in it are still less than o.i per cent. Potentiometer Reading, 0.011241; Temperature, 229.3. Magnet Current. Bridge Reading. Resistance of Film. dr dr/r H 20.957 20.587 1.0 20.943 20.573 20.930 20.560 0.013 0.0005 2300 2.0 20.965 20.595 20.938 20.568 0.027 0.0013 4725 3.0 21.001 20.631 20.930 20.560 0.071 0.0034 7250 4.0 21.039 20.669 20.930 20.560 0.109 0.0053 9700 5.0 21.086 20.716 "' 20.925 20.555 0.161 0.0078 11950 6.0 21.141 20.771 20.923 20.553 0.218 0.0106 14125 6.8 21.186 20.816 20.921 20.551 0.265 0.0129 15650 For measurements at low temperatures the oil bath was removed from between the pole-pieces of the magnet, and, to make possible better heat insulation, the pole-pieces were separated to a distance of 1.8 cm. The exposed parts of the magnet in between the coils were covered with a thick layer of heat insulation, a thinner layer also covering the faces of the pole-pieces. A cardboard container for the refrigerant was lowered into the rectangular compartment thus formed. A vertical section is shown in Fig. 3 which will give an idea of the arrangement. Since the pole-pieces of the magnet had been separated to a greater distance it was necessary to increase the current through the coils, beyond that which was used during the experiments at the higher tem- peratures, in order to cover the same range of field strengths. The magnet was accordingly connected to the no-volt direct current supply of the laboratory through suitable rheostats. This arrangement was not as satisfactory as the storage battery had been since the fluctuations 262 L. F. CURTISS. [SECOND LSERIES. Fig. 3. in the line voltage caused small variations in the magnet current. How- ever, by exercising extreme care, errors from this source could be reduced to a negligible amount. The maximum current used was about 18 amperes, giving a field of about 18,000 gauss. The magnet was recali- brated, a ballistic galvanometer, a standard of mutual in- ductance, and a snatch coil being used for this purpose, since an accident to the bismuth spiral had ruined it. The films were removed from their protecting glass tubes for the measurements at low temperatures, as they have very slight tendency to oxidize at room temperature and below, and a considerable saving of space between the pole-pieces was thereby accomplished . They were mounted on a fiber support by means of which they were accu- rately centered between the pole-pieces. One junction of a thermo-couple for measuring temperatures was also mounted on the fiber strip so that it was close to the film. Great difficulty, due to poor heat insulation and the conductivity of the large pole-pieces, was experienced in maintaining constant temperatures for any considerable period of time unless the films were actually immersed in the refrigerant. Consequently measurements in the field were only attempted when the films were immersed in either a mixture of carbon dioxide snow and ether or in liquid air. The results of the investigation can be considered under four heads, viz., (i) effect of magnetic field on the resistance of the films at various temperatures; (2) the dependence of the values of the percentage increase of resistance upon the thickness of the film; (3) evidences of aging of the films; (4) relation between the temperature of the film and its resistance. Observations were made on the behavior of a large number of films varying in resistance from 7 ohms to 70 ohms. No attempts were made to measure the thickness of the films. Since all of the films had the same length and breadth, the resistance can be taken as a rough indication of the thickness. Film V-io has been selected as an example for a discussion of the effect of temperature upon the percentage increase of resistance of the films in a magnetic field. At 20.5 its resistance was 10.54 ohms, hence it was one of the thicker films studied. The results of the measurements on this film at various temperatures are shown graphically in Fig. 4. The percentage increase of resistance has been plotted against the field strength in kilogauss. The various curves have been labelled and were taken at the following temperatures: PHYSICAL PROPERTIES OF THIN METALLIC FILMS. 26 3 I. II ill 20.5 C 57.2 C 109.0 C IV 147.5' V 195 VI.. ..227 VII - 80 VIII - 190 A.. 19.5< As the temperature decreases the effect of the field becomes more strongly marked, especially at the low temperatures. For the weaker fields there is a decided curvature in the lines which gradually smooth z >^ fi^ ^7 ^ y 2J* *> w xw /EJ* /so- ny zoo" & Temperature Fig. 9. resistance measurements here shown were undertaken, and that in making the first heating the previous maximum temperature to which the film had been heated was exceeded with a resultant decrease in resistance Film V-/3 o - First Htulina & - Second CW> >- ' Ceofinj -1ff4gHgy Second /fc*rm s +- " c <" i \ \ \ I \ 's \ 9 \ \ \ A \ ^ ^ i , , Tcm i >e r a fj /re ^r- i. 10. VOL. XVIII. No. 4 ] PHYSICAL PROPERTIES OF THIN METALLIC FILMS. 269 of the film. One would expect then that this would show itself in the behavior of the film in the magnetic field. Data taken in the magnetic field before and after these temperature-resistance measurements yielded * Film V-5 - FirT H ti nj - S ecortd C.o/my ^-Second ' o-TA.rd Hertinj A-rThird Cooling 20.0 195 \ / I \ 1 \ / 110 \ \ / . "emperalure Fig. 11. the expected result. Since too many figures would be required to show all the results in detail, an idea of the relative magnitude of this change produced in the values of dr/r may be obtained by reference to Fig. 4, IDA 9.JS Y v Film 0-finT Heating 0- " Cooling V-/o - Second He.f,f A- " Cooling Temperature Fig. 12. The above figures show typical temperature-resistance curves for films of different thickness. 2 7 L - F - CURTISS. where the similar results for Film V-io are represented. The curves labelled / and A show the behavior of this film when treated in a like manner. Curve / was taken before the temperature-resistance run and Curve A afterwards at the same temperature. The result of the aging was, as shown, to increase the values of dr/r. A further point of interest in Fig. 9 is the suggestion of a minimum value of the resistance at about 210. This minimum shows much more clearly in the curve for Film V 13 in Fig. 10. It occurs, moreover, at a lower temperature, in the neighborhood of 170. In this connection it is to be noted that -13 has a considerably lower resistance than the preceding film. Further- more the preliminary heat treatment was carried out much more thor- oughly, the results of which are strikingly shown by the reproducibility of the data. The observed points are here indicated by various symbols for three heatings and three coolings, taken over a period of several days, and yet a single curve represents all fairly well; only the very slightest traces of aging are to be detected. Likewise it was found possible to reproduce the measurements in the magnetic field with similar accuracy for this film. These results are quite contrary to any of the author's previous experiences with sputtered films. In fact, it had always been found impossible to so age a film that its resistance would remain con- stant for more than a few hours at room temperature, while exposed to the air, and reproducible measurements at higher temperatures were out of the question. To show more completely the form of the tempera- ture-resistance curve the data for Film V~5 are shown in Fig. 1 1 . Here both branches of the curve, which is perfectly symmetrical and in form approaches that of a parabola, are well defined. This film also has been thoroughly aged and the minimum resistance has shifted to a still lower temperature. An attempt to fit a parabolic equation to this curve has been made. Although the deviations are slight, they seem to be some- what greater than possible experimental errors. All previous films of bismuth which have been studied by the author, or by others, so far as a search of the literature has revealed, have without exception had negative temperature coefficients of resistance. We have here, how- ever, an example where the coefficient, initially negative at room temperature, actually reverses its sign as the film is heated and be- comes positive. The maximum value of the positive coefficient which has been obtained is far less than that for bismuth in bulk, roughly of one quarter the magnitude. A further example of this form of curve is shown in Fig. 12, representing the results obtained with Film V-io. This is another example of a film that has been only partially aged, hence there was a big decrease in resistance as the film was heated for PHYSICAL PROPERTIES OF THIN METALLIC FILMS. 2JI the first time. The effect of the second heating, as was to be expected, was considerably less, although it was carried out to a slightly higher temperature. The symmetry of the curves is considerably distorted by this aging effect which takes place at the higher temperatures. It is to be noted that the points representing the second heating lie closely on the curve for the first cooling, until the higher temperatures are reached, showing that the aging occurs chiefly at temperatures above 200. \ nim v-/t. \ \ \ \ \ \ \ \ s ^ r^~ Fig. 13. Showing the complete temperature resistance-curve for Film V-io. In Fig. 13 the complete temperature-resistance curve for Film V 9 is represented. The part from room temperature to 230 was taken some months before the part at low temperatures. This fact, together with the effect of increased pressure to which the film was subjected when air was admitted to it, accounts for the jog in the curve at o. If the part taken at the higher temperatures were raised throughout by a constant amount, equal to the difference of ordinates at o, the two parts would then form a smooth curve which agrees in form with those taken for other films throughout this range of temperatures. Hence there seems to be no evidence of any peculiarities in the low temperature part of the temperature-resistance curves for these films. To show what happens to one of these films when it is heated in contact with the air, a temperature-resistance run for Film V-8 is shown in Fig. 14. This film was prepared in the usual way and during the first heating and cooling behaved much the same as had all previous films which were thoroughly aged, the two sets of observations lying along the 272 L. F. CURTISS. SECOND SERIES. same curve. However, soon after the beginning of the second heating, the resistance began to increase rapidly. After cooling down it had a considerably higher resistance than before at room temperature. This anomalous behavior was hard to explain until the film was taken out of the bath and examined. It was then discovered that a yellowish coating of oxide had form on the surface, indicating that the glass tube had given 29 \ V Film V- O-fivT Coolint - Second He* ling TemperaTure Fig. 14. This figure shows the temperature-resistance curve obtained for Film V-8, which was attacked by the air as a result of a break in the protecting tube. way at some point, admitting air. On further examination this proved to be the case. There seems to be an additional effect of the air on the form of this curve beyond that produced by oxidation alone, however. The difference in ordinates at the maximum temperature is considerably greater than at room temperature, hence the increase of resistance at the higher temperature is not all permanent, as it would be if it were due solely to oxidation. This additional effect may well be caused by the pressure, although this point was not specially investigated. The foregoing covers the experimental results of this investigation. As far as the magnetic measurements are concerned little more can be said beyond the mere statement that these films approach the behavior of bismuth in bulk, the approach being nearer the thicker the film. Any theoretical discussion of these results and interpretation in terms of the electron theory would necessarily follow the same general lines as those for bismuth in bulk. Too little is known at present with any degree of certainty about the nature of electrical conduction in metals to enable very much progress to be made in this direction beyond what has already been accomplished by Thomson and others. This is very clear when PHYSICAL PROPERTIES OF THIN METALLIC FILMS. 2 73 attention is called to the fact that the most successful theories so far advanced, 1 either agree with the experimental data only for very limited ranges of the field strength, or require properties which bismuth does not possess, e.g., magnetostriction. It is also difficult to explain successfully the form of the temperature- resistance curves of these films. One explanation which might be offered for the reversal of the sign of the temperature coefficient and the consequent minimum resistance, the location of which varies from film to film, and is, in general, at a lower temperature the lower the resistance of the film, is as follows: It is based on the fact that the metal particles have a larger coefficient of thermal expansion than the glass. Hence as the film rises in temperature they expand more rapidly and bring more particles into contact, reducing the negative temperature coef- ficient of resistance, and finally, if the film is thick enough, the particles are crowded close enough together and at a rate fast enough to reduce this coefficient to zero, then finally to reverse its sign. Then the metallic film begins to behave somewhat like the metal in bulk, but is prevented from attaining a positive coefficient as large as that for solid bismuth by the fact that the glass also continues to expand, thus preventing the compacting process from going far enough to establish contact between all the possible particles. The study of this phase of the subject is limited by the comparatively low melting point of bismuth (about 260). This explanation is far from satisfactory, however, since on the basis of it a platinum film sputtered on glass should have practically a zero tempera- ture coefficient, which is contrary to the facts as reported by other observers. But then there must be taken into account also the effect of temperature on the conductivity of the metallic particles themselves. Hence we have at least two influences which must be considered simul- taneously. The first is that due to the difference in the thermal coef- ficients of expansion which should cause an increase or decrease in the number of possible paths for electrons from particle to particle, according as the film is heated or cooled. The second is that which results from the effect of temperature upon the motions of the electrons within the particles, and also its effect upon the ease with which an electron may detach itself from one particle and move on to the next. This view of the structure of the films is strengthened by the fact that x-ray studies 2 of the films indicated that they were composed of small crystals of metal, with random orientation, which have been detached from the cathode and deposited on the glass during sputtering. This view is in contrast 1 E. P. Adams, PHYS. REV., 24, p. 248, 1907. 2 H. Kahler, PHYS. REV., 17, p. 230, 1921. 74 *- F - CURTISS. rith the evaporation theory which holds that the metal of the cathode > vaporized and then condenses on the glass, in which case a film of ntirely different structure from that of the cathode metal might result, lowever, this point is not settled. For further comparison of the results here given with those obtained 3r bismuth in bulk, a reference to an investigation by Dr. F. C. Blake 1 > included. His work, published in 1909, covers practically the same ange for the solid metal. SUMMARY. The preceding results may be summarized as follows : 1. By protecting sputtered bismuth films in a vacuum and heating hem to temperatures near the melting point of bismuth it is possible to evelop the property of an increase of resistance in a magnetic field to bout one fifth that possessed by bismuth in bulk. 2. The higher the temperature of the film the smaller the value of the ercentage increase of resistance for a given field. 3. The form of the curve between dr/r and field strength is independent f the temperature of these films, i.e., dr/r = kf(H). 4. The temperature-resistance curves for these films are parabolic ti form, and the minimum resistance is at a lower temperature the thicker he film. 5. If these films have been subjected to a very thorough heat treatment he temperature-resistance curve is fixed and definite. Under these onditions reproducible data are easily obtained in the magnetic field. 6. If the film is only partially aged it decreases in resistance with con- inued heating and exhibits correspondingly greater values of dr/r in the nagnetic field as a result of the heating. In conclusion I wish to thank Professors Ernest Merritt and F. K. lichtmyer, under whose direction this work was done, and all others, yho, by advice or suggestions, have helped the progress of this investi- gation. PHYSICAL LABORATORY, CORNELL UNIVERSITY. 1 F. C. Blake, Ann. d. Phys., 28, p. 449, 1909. THE AMERICAN PHYSICAL SOCIETY OFFICERS OF THE SOCIETY President: THEODORE LYMAN Harvard University, Cambridge, Mass. Vice-President: C. E. MENDENHALL Univ. of Wisconsin, Madison, Wis. Secretary: D. C. MILLER Case School of Applied Science, Cleveland, Ohio Treasurer: G. B. PEGRAM Columbia University, New York City Editor: F. BEDELL Cornell University, Ithaca, N. Y. Local Secretary for the Pacific Coast: E. P. LEWIS University of California, Berkeley, Cal. The Council of the Society consists of the president, vice-president, secretary, treasurer, manag- ing editor, all living past presidents and eight elected members as follows: Past Presidents: A. A. Michelson, A. G. Webster, Carl Barus, E. L. Nichols, Henry Crew W. F. Magie, Ernest Merritt, R. A. Millikan, J. S. Ames. Elected Members: A. L. Day, G. F. Hull, G. K. Burgess, J. C. McLennan, F. B. Jewett, Max Mason, K. T. Compton, H. M. Randall. A complete list of past officers is given on a back page of the REVIEW. THE PHYSICAL REVIEW Published Monthly Two Volumes Annually Annual Subscription, throughout the Postal Union, $7.00, in advance; Single Copies, $.75 General Index 1893-1920 (issued June, 1921) $4.00, postage extra. Back Volumes (either series) $4.50 per volume, carriage extra. Back Numbers (either series) $-75 a copy, postage extra. Discount to subscription agencies only, ten per cent. NOTE. When remittance accompanies the order, the Index will be sent prepaid for $4.00 and single back numbers will be sent prepaid for $.75. Back numbers will only be sent when they can be furnished without breaking a volume, and back volumes when they can be fur- nished without breaking a continuous series. Orders that can not be completely filled will be filled so far as stock permits, unless otherwise specified in the order. Money will be refunded for any defective copies returned. No bound volumes can be supplied. The REVIEW offers to purchase, at $1.50 a copy, a limited number of No. i, Vol. I, First Series; and No. 2, Vol. IX, February 1917, Second Series. Notaries certificates will be furnished with bills, when so ordered, for twenty-five cents each; copies of bills, on forms furnished by customer, will be furnished for five cents each. Stamps will be accepted for these. Members of the Physical Society may obtain back numbers at a discount of 10 per cent, provided statement of membership accompanies order. When a member's name has been dropped from the mailing list of the REVIEW on account of non-payment of dues, it will be restored to the list when dues are paid; any numbers that he has failed to receive in the interim will be furnished without cost upon application to the REVIEW. All complaints for failure to receive the Review should be addressed to the Review and not to any officer of the Society. First Series. From July i, 1893, to January i, 1913, the REVIEW was conducted by Edward L. Nichols, Ernest Merritt and Frederick Bedell, with the financial support of Cornell Uni- versity. During this period thirty-five volumes were issued, constituting the First Series. The first four volumes were issued annually; subsequent volumes, semi-annually. Second Series. On January first, 1913, the REVIEW was transferred to the American Physical Society. A second series was then begun, the January, 1913, number being desig- nated Volume I., Number I, Second Series. (An announcement of this transfer will be found on the first page of the new series.) Semi-annual volumes begin in January and July. For advertising rates, address THE PHYSICAL REVIEW, Ithaca, N. Y. Change of copy in standing advertisements should be sent direct to the printers, THE NEW ERA PRINTING Co. Lancaster, Pa. Old copy is used until new copy is received. Contributors to the REVIEW are referred to page two of the advertisements for further information. Permission is given to republish any article appearing in the REVIEW, provided proper credit is given. Address Correspondence to THE PHYSICAL REVIEW, ITHACA, N. Y. THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW AN INITIAL FINE OF 25 CENTS WILL BE ASSESSED FOR FAILURE TO RETURN THIS BOOK ON THE DATE DUE. THE PENALTY WILL INCREASE TO SO CENTS ON THE FOURTH DAY AND TO $1.OO ON THE SEVENTH DAY OVERDUE. JAM 11 1935 LD 21-100m-8,'34 UNIVERSITY OF CALIFORNIA LIBRARY