The excitation of y-radiationin processes of proton capture

In this paper the results of an investigation of the 7 -radiations emitted in the process of proton capture by the four light elements, beryllium, boron, carbon and fluorine, are presented. The intensities of the capture radiation for different energies of the incident protons have been determined from the numbers of coincidences produced in a pair of Geiger-Mueller counters. The high sensitivity and low natural effect of this arrangement render it very suitable for the detection of resonance effects of low intensity. Moreover, such weak resonances could be separated from more pronounced ones by the use of very thin targets from which the yield of y-radiation was relatively low. By this method the excitation curves of the y-ray emission have been determined up to an energy of the incident protons of 102 *6 eV, and new resonances have been found for each of the above elements. Measurements of the quantum energy of the emitted y-radiation have been made for each of the four elements. I t has been possible, by analysis of the experimental data, to distinguish between the effects of proton capture by the constituent isotopes in the case of boron and of carbon. In the case of carbon the yield of y-quanta per incident proton has been found for both isotopes. T h e ex c ita tio n o f 7-rad iation in processes o f proton capture by ligh t e lem en ts


I ntroduction
In this paper the results of an investigation of the 7 -radiations emitted in the process of proton capture by the four light elements, beryllium, boron, carbon and fluorine, are presented.
The intensities of the capture radiation for different energies of the incident protons have been determined from the numbers of coincidences produced in a pair of Geiger-Mueller counters. The high sensitivity and low natural effect of this arrangement render it very suitable for the detection of resonance effects of low intensity. Moreover, such weak resonances could be separated from more pronounced ones by the use of very thin targets from which the yield of y-radiation was relatively low. By this method the excitation curves of the y-ray emission have been determined up to an energy of the incident protons of 102 * * * 6 eV, and new resonances have been found for each of the above elements.
Measurements of the quantum energy of the emitted y-radiation have been made for each of the four elements. I t has been possible, by analysis of the experimental data, to distinguish between the effects of proton capture by the constituent isotopes in the case of boron and of carbon.
In the case of carbon the yield of y-quanta per incident proton has been found for both isotopes.

T h e e x c ita tio n o f 7-r a d ia tio n in p r o cesses o f p roton c a p tu re b y lig h t e le m en ts 2. The ex per im en ta l arrangem ent a n d detecting apparatus
The experiments here described were carried out in the Cavendish High Voltage Laboratory,.using the Philips generator and the associated dis charge tube installation from which beams of ions with energies up to l*2x 106 eV are available. Magnetic analysis of the accelerated ion beam was made, and the experiments are concerned only with the effects of proton bombardment.

S. C. Curran, P. I. Dee and V. Petr&ilka
The brass target tube T ( fig. 1) contained a rotatable, water-cooled copper block B having two flat faces, either of which could be used to support targets in a plane perpendicular to the path of the beam. Usually, however, one of these two faces was used as a blank target for control experiments. The proton current incident upon the target ranged between 30 and IOO/a A, being in most cases about 70 The variation of the current in the course of an experiment was less than 2 .%.
For the detection of the emitted y-rays two thin-walled copper or brass Geiger-Mueller counters (C1 and C2) were employed in coincidence. The diameter of the tubes was 2 cm. and the distance of separation of their axes was 4 cm. The counters were placed inside a lead box with walls several cm. thick. An aperture A in this box permitted the passage of the capture y-rays to the counters.
A type of internally extinguished counter with thin walls was developed for use in these experiments. The thin-walled part of the tubes was about 6 cm. long and 0T mm. thick. The metal tubes were of copper or brass. The filling gas consisted of argon (partial pressure 9 cm.) and alcohol (partial pressure 1 cm.) (Trost 1935a(Trost , b, 19361937). More details concerning these counters will be published elsewhere. These counters had a flat portion on their characteristic curves (counts x volts) of some 200 V.
A resistance of 1 MU was connected in series with each of the counters, and the voltage pulse developed across this resistance was applied to the grid of the first valve of a two-stage resistance-capacity coupled amplifier. The amplified pulses from the two counters were mixed in a Rossi (1932) circuit ( fig. 2 inset), and coincidences were counted by a three-stage "scaleof-two" counting meter (Wynn Williams 1932;Lewis 1937).
Various factors influence the behaviour of the tubes in such a circuit. If Nt counts/min. are produced in one counter and N2 counts/min. are produced in the other counter, a number min. of chance coincidences will be recorded. N0, Nx and N2 are connected by the equation T = 602^/221 sec., where T is known as the " resolving tim e" of the counting system. I t was desirable to reduce the number of chance coincidences N0 to a minimum. W ith our experimental arrangements there were two means of achieving this result. When the two counters were separated and screened from one another and two independent sources of /^-radiation were used to of y-radiation in processes of proton capture 271 irradiate them so th at their individual rates of counting were Nt and NJmin., a number NJmin. of chance coincidences was recorded when the were working in coincidence. For a fixed value of NXN2 it was found that N0 increased rapidly with the voltage applied to the tubes as shown by curve 2 of fig. 2. If, however, the counters were placed close together and a source of /^-radiation was placed so th at the /?-rays could pass through both tubes, producing real coincidences, the number of such real coin cidences which was recorded for different voltages on the pair of tubes was found to be as shown in curve 1 of fig. 2. With a fixed strength of /?-ray source the number of real coincidences per minute was constant over a large range of voltage. In our experiments, therefore, the counters were operated a t voltages which corresponded to the lower end of the flat portion of curve 1, fig. 2. It was also found that the resistance in series with each of the counters exerted a marked influence on the number of chance coincidences (NJmin.) which was recorded for a fixed value of Nj_N2. With our arrangement the optimum value of this resistance was found to be 1 Mi2. In addition to these two considerations a slight reduction in T could be obtained by suitable adjustment of discriminator bias. A resolving time of 10-5 sec. was finally obtainable with this simple and reliable arrangement.
We have used this counting system for the detection of the -/-radiation and for the measurements of its quantum energy. The capture radiation pro duced at the target B ( fig. 1) and falling upon the aluminium plate P which closed the aperture A, ejected Compton electrons, photoelectrons and elec tron pairs from this plate. Some of these passed through both counters causing real coincidences. The range distribution of these electrons was deter mined by placing absorbing screens of different thicknesses between the counters and observing the number of coincidences as a function of the thickness of absorber. Information concerning the quantum energies of incident y-radiations may be derived from such absorption curves in the following manner. The end-point of the absorption curve indicates the maximum range of the electrons and hence gives a measure of their limiting energy (Rasetti 1935). In the absence of precise data relating range and energy for electrons with energies of the order of 107 eV, it is necessary to obtain empirically a calibration curve connecting the observed end-point with the maximum quantum energy of the incident radiation. The curve which we have used for this purpose is shown in fig. 10, and an account of the method by which this curve has been obtained is given in § 5. Informa tion concerning the quantum energy of the incident y-radiation may also be derived from the value of absorber thickness at which the rate of coin cidence counting falls to one-half of its value for no absorber between the counters (Fleischmann 1936). Mitchell and Langer (1937) and Kikuchi (1936aKikuchi ( , b, c, 1937 have published data relating this half-value thickness with the quantum energy of the y-radiation. I t appears, however, th a t considerable divergence exists between the results of those who employ this method. A knowledge of this half-value thickness may often, however, be combined with a knowledge of the end-point to provide some information concerning the homogeneity or inhomogeneity of the incident radiation. For the study of radioactive effects produced by the capture of protons the target tube was provided with a window ( fig. 1) of aluminium foil 0-1 mm. in thickness.
The excitation of y-radiation in processes of proton capture 273 3. P reparation of the targets Carbon targets were obtained either by deposition of soot from a candle flame or by evaporation of aquadag. Both methods yielded targets of a uniform thickness which could be adjusted a t will. The metallic beryllium used in the preparation of Be targets was an H.S. brand pure substance (obtained from Messrs Hilger, Ltd., Lab. No. 7778), spectroscopic examina tion of which had shown it to be free from lithium and fluorine. The beryllium targets were obtained as thin bright metallic layers upon copper plates by evaporation of this metal in vacuo. Very different thicknesses of target substance could be deposited in this manner. In the case of boron the same method was employed. Because the impurity content of the amorphous boron was .not known, analytically pure B20 3 (Schering and Kahlbaum) was also used. No differences between the results for these two targets could be found. For the experiments with fluorine, BaF2 (Schering and Kahlbaum) was evaporated on to the target block in air. Layers which were hardly visible were found to give sufficient intensity of y-radiation for the purposes of the investigation and to remain unaltered during the course of an experiment.

(a) Beryllium
The excitation curve of the y-radiation from beryllium under bombard ment with protons has been examined by two groups of workers. Hafstad and Tuve (1935) first detected the capture of y-radiation from this element using a thick target. Their results extended from about 600 kV (below which they could not detect radiation in sufficient intensity for accurate measure ment) to 1000 kV, and they suggested th a t the process was not a resonance one. In later work, Herb, Kerst and McKibben (1937) extended the results beyond 1000 kV, and showed th at a broad resonance was present a t about 1 mV. W ith our more sensitive detecting apparatus we have extended the study of the y-radiation emitted in this process down to a voltage of 200 kV. We have examined, using pure Be targets of varying thicknesses, the course of the excitation curve from 200 to 1000 kV. The relationship between y-ray intensity and bombarding voltage for a target of Be metal is shown in fig. 3, curve 1. The thickness of the target in this case is estimated as equal to the range of a proton of 20,000 eV energy. Our results are in rough agree ment with those of Hafstad and Tuve for the region above 600 kV. Curve 2 of fig. 3 shows the results of a more detailed investigation of the region from 200 to 600 kV. I t may be seen th a t a resonance peak is present at a bom barding voltage of 350 kV. The small peak at 480 kV and the rise of intensity at about 550 kV may be attributed to contamination of the target by carbon deposited during the bombardment (see § 4 A similar investigation of thelregion from 550 to 850 kV revealed a second new resonance with a peak intensity at 670 kV. This region is shown in more detail as curve 3 of fig. 3. The ion source used in these experiments was of the discharge tube type and an inhomogeneity of about 20,000 eV may exist in the energy of the protons from such a source. This inhomogeneity of the ion beam makes it impossible to draw any precise conclusions concerning the widths of the resonances which have been observed. It appears, however, from these results th at the widths of the two new resonances at 350 and 670 kV, due to the bombardment of beryllium with protons, are considerably greater than the widths of the resonances associated with the bombardment of fluorine (see § 4 (e)).
The quantum energy of the capture radiation emitted in the process was investigated by Crane, Delsasso, Fowler and Lauritsen (1935 a). Expan sion chamber photographs of the tracks of electrons projected by the capture radiation gave evidence of the presence of quanta with energies 2*2, 3*7, 4-8 and 6-0 x 106 eV. For zero bombarding energy of the incident proton the maximum quantum energy available in process (1) is, by calculation from accepted mass values, 6-39 x 106 eV. The absorption curves which we have obtained for this element for different energies of the incident protons are shown in fig. 4. I t may be seen from the results th at the end points, according to our calibration curve, give values for the maximum quantum energy in good agreement with the values calculated from equation (1), when allowance is made for the different energies of the incident protons. These results prove th at the radiation is, a t least to a very great extent, due to the process of proton capture. In the case of beryllium bombarded with protons only one other process might lead to the emission of y-radiation of appreciable energy: 9Be + X H -> 6Li + 4He.
A maximum quantum energy of about 2 x 106 eV would be possible from this reaction, but no excited states of 6Li or of 4He are known at such low energies. Moreover, the form of the absorption curves clearly shows th at the bulk of the radiation is very much harder than 2 x 106 eV. Thus the half-value absorption thickness indicates, using the relationship due to Mitchell and Langer (1937), a y-ray energy of 5*0 x 106 eV a t 400 kV, and an energy of 5*7 x 10® eV at 850 kV. These results, while in accordance with the known inhomogeneity of the radiation, prove th at most of the radiation arises from the process (1). With the help of these facts concerning the observed energy release it is interesting to consider further the form of the excitation curve of the y-radiation. I t may be seen from the curve th at the two resonances at 350 and 670 kV are weak effects superimposed upon a general background of radiation. The form of the curve is roughly exponential, the intensity of the background increasing smoothly with bombarding voltage, and it appears from the absorption measurements th at this background radiation mu t be associated with the capture process (equation (1)). The excitation The excitation of y-radiation in processes of proton capture 275 curve may then be regarded as due to the existence of a broad excited state in the resultant 10B nucleus, extending between energy values of about 6-4 and > 7-4 x 10* 6 7 eV above the ground state, the weak resonance effects at 350 and 670 kV being due to excited states of considerably smaller widths at energies 6-75 and 7-07 x 106 eV above the ground state. In connexion with this possibility of the existence of a broad state of excitation of the 10B nucleus it is interesting to notice th a t Gentner (1937) has found th a t the emission of y-radiation from lithium under proton bombardment is not entirely confined to the regions of sharp resonance, a small but definite y-ray emission being observed for all energies of the incident protons. I t should be noted, however, th a t in the latter case the quantum energy is of the same magnitude as the energy release in the process of disintegration into a-particles, and the possibility of the broad resonance being due to the formation of a-particles in excited states has not therefore been excluded. We have given reasons for believing th a t the broad resonance observed in the case of beryllium cannot be attributed to excitation of the products of the reaction.
Alternatively, it is possible th a t the exponential form of the excitation curve is due to a lack of resolving power in the present experiments. I t is possible, for example, th a t a large number of resonances exist in the region from 0 to 1 mV, the intensities of these resonances increasing with voltage and combining to produce an apparently exponential curve. On either view, the small changes of the maximum quantum energy which have been observed to be produced by changes of the bombarding voltage are to be expected.
(6) Boron The excitation curve for the y-rays emitted in the process n B + HI -> 12C + h v ( 3) has been studied by Herb, K erst and McKibben (1937), using a thick boron target of unknown purity. They obtained a smooth increase in intensity of y-radiation with voltage, except for a weak resonance a t about 820 kV. Bothe and Gentner (1937®) investigated the excitation at lower voltages from 100 to 500 kV. They found a resonance effect at 180 kV and a second rise at 360 kV, which they attributed to the effect of the molecular beam. They showed th at the excitation curve of the y-radiation ran parallel to the excitation curve of the a-particles produced in the process which was studied by Williams and Wells (1936) and by Williams, Wells, Tate and Hill (1937)-Since it seemed desirable to obtain more accurate data on the excitation function between 300 kV and 1 mV we have investigated this region of voltage with great care. The results are shown in fig. 5. Curves 1 and 2 are the results obtained with thin targets of boron and boron trioxide.

e excitation of y -r a d i a t i o ni n processes of proton capture 277
Vol. CLXIX. A.

S. C. Curran, P. I. Dee and V. PetrZflka
The relative yields from these two targets may not be taken from the figure, since conditions of measurement were not the same in both cases. Curve 3 gives the measurements between 300 and 600 kV, using a B20 3 target, in greater detail. No evidence of any resonance effects may be derived from this curve. This confirms th at the effect which Bothe and 900 kV Gentner (1937a) found at 360 kV was due to the molecular beam as they suggested. I t should be noted, however, th at the y-ray intensity as shown by this curve exhibits a fourfold increase between 300 and 500 kV, whereas Bothe and Gentner found an increase of only 70 % in this region. From 300 to about 800 kV the y-ray intensity increases smoothly with voltage, but at 850 kV reaches a first resonance peak followed by a second one at 950 kV. There is also slight evidence of a change of gradient at about 650 kV, and this might indicate a third resonance.

The excitation of y-radiation in processes of proton capture 279
Several energy measurements for the y-rays emitted in process (3) have been carried out. The preliminary work of Crane, Delsasso, Fowler and Lauritsen (19356) indicated a maximum quantum energy of 14*5 x 10® eV. Bothe and Gentner (1937 a) found a maximum quantum energy of 14 x 10® eV which was later re-determined by Gentner (1937) as 13 x 10® eV.
The maximum quantum energy to be expected by substitution of the mass values in process (3) is 16*67 x 10® eV a t a bombarding energy of 850 kV, and it was generally supposed th a t the resultant 12C nucleus was left in an excited state of about 3 x 10® eV.
A detailed measurement of the y-ray energy spectrum by Fowler, Gaerttner and Lauritsen (1938) has revealed the presence of three lines of energies 4*3, 11*8 and 16*6 x 10® eV. I t seems certain from these results th a t y-rays with quantum energies corresponding to the maximum energy release indicated by equation (3) are emitted.
Measurements of the end-point of the absorption curve have given the value 28*0 mm. of aluminium at a bombarding voltage of 850 kV, and we have taken this value as corresponding to a quantum energy of 16*67 x 10® eV in the construction of the curve relating absorption thickness with quantum energy, (see § 5).

(c) Radioactive effects from boron
We have discussed above the capture process for boron without reference to capture by the isotope 10B.
The formation of a radioactive positron emitter with half-life 20*5 min. has been reported by Crane andLauritsen (1934a, 1935) to result from the bombardment of boron with protons having an energy of 900,000 eV. The formation of this radioactive body was attributed to the process 10B + -* UC + h v .( 5) The quantum energy of the y-radiation to be expected from this process is 9*4 x 10® eV, but no evidence of this energy value was found in the measure ments of the y-ray spectrum which were made by Fowler, Gaerttner and Lauritsen (1938). These latter workers considered therefore th at the absence of any y-rays of this energy might be explained by assuming th at the pro duction of radioactive UC was a resonance process occurring at a voltage of about 900 kV, a somewhat higher value than that used in their own experi ments. Some support for this view appeared to be given by the fact th a t Allison (1936) found no evidence of the formation of a radioactive body with this period for proton energies as high as 500,000 eV.
In a series of experiments with proton energies varying from 350,000 to 750,000 eV we have been unable to find any definite evidence of the formation of a radioactive body with a half-life of 20-5 min. Since it appeared th at in these experiments radioactive effects were being produced by the action of the y-rays upon the target tube, etc., further experiments were made to eliminate such effects. A thick target of boron was bombarded with 60 nA of protons of energy 960,000 eV for 30 min., the target w withdrawn from the tube and the boron scraped away from the target holder. Analysis of the decay curve of the radioactivity of this boron showed the presence of two periods of values 10-7 min. and ~ 40 days. The most natural interpretation of this result is to associate the 10-7 min. period with radioactive 13N formed by bombardment of 12C present as contamination of the target, while the 40-day period may be attributed to radioactive 7Be (half-period 43 days) formed as a result of the process 10B + 1H -» 7Be + 4He (Roberts, Heydenburg and Locher 1938). Similar experiments carried out at different bombarding voltages yielded the same results. According to our measurements the yield of positron emission, having a half-life 20*5 min., is certainly less than 10~13 per proton, whereas the yield according to Crane and Lauritsen (1935) was stated to be about 10~10 to 10-11 per proton. I t is possible th at the results of Crane and Lauritsen were due to photo disintegration effects, produced by the y-rays from n B, since we found evidence of periods 10-5 and 38 min. in the earlier experiments in which the targets were not removed from the target tube. These periods could be attributed to the action of the y-radiation upon the copper and zinc present in the brass of the target tube (Bothe and Gentner 19376).
As a final test of the correctness of the above conclusions a boron target was bombarded with 50 juA of deuterons at 750 kV for 10 min. Analysis of the decay curve of the intense radioactivity which was produced in this case gave clear evidence of a period 21 min. which may be attributed to n C formed according to the process 10B + 2H -> n C + In. From the yield of the radioactive effect which was produced in this case it was clear th at the amount of heavy hydrogen present in natural hydrogen would be sufficient to produce this activity in measurable amount. This observation may also be advanced as a possible explanation of the effect observed by Crane and Lauritsen who did not use magnetic separation of the bombarding beam.

.(d) Carbon
The capture of protons by carbon according to the process was investigated by Hafstad and Tuve (1935) who measured the amount of radioactive 13N which was formed at different bombarding voltages. These workers detected the emission of y-radiation but were unable to determine its excitation curve owing to the low sensitivity of their detecting apparatus. We have used counters in coincidence for the study of this capture radiation with the result shown in fig. 6, curve 1. Our earlier measurements (Dee, Curran and Petrzilka 1938) have been extended to 950 kV. Two resonance effects a t 480 and 570 kV are Evident. We have also extended our earlier measure ments of the positron emission from 13N. This was done by bombarding the target for 3 min. a t different voltages and measuring the subsequent acti vities. The fact th a t 13N was formed was checked by evaluation of the period which was found to be 9*9 min. This agrees satisfactorily with the accepted value of the period making allowance for the escape of 13N from the target, which was in vacuum during measurement. The results are shown in curve 2 of fig. 6. I t is clear th a t only the lower resonance of curve 1 should be associated with the process (6). I t is natural to associate the resonance at 570 kV with the process 13C + 1H -> 14N + ^, The excitation of y-radiation in processes of proton capture 281 in which capture of protons leads to the formation of stable 14N. Consideration of the energy balance in these two reactions indicates th a t process (6) results in the release of y-rays of quantum energy 2-37 x 10® eV at a proton bombarding voltage of 470 kV, while for the process (7) the energy of the emitted y-ray quantum is 8*23 x 106 eV a t 570 kV. Measure ments of the y-ray energies a t 470 and 560 kV yielded the results shown in fig. 7. Our earlier work had indicated a y-ray energy of 7*4 x 106eV at560kV, but further measurements with an improved counting system, a t a bombardingvoltageof 600 kV, indicated a y-ray energy of 8*5 x 106eV (curve 1). This value of the quantum energy was obtained from the measured end-point value of curve 1. Consideration of the ' 'half-value ' ' absorber thickness indi cates a quantum energy of about 6 x 106 eV. This fact may be explained by the assumption th a t the excited 14N nucleus often reverts to the ground state by the successive emission of two quanta each of energy ~ 4 x 10® eV (Rose 1938). This is in good agreement with the quantum energy 8*2 x 106 eV expected from process (7). The energy of the y-rays was also measured with a bombarding voltage of 480 kV (curve 2, fig. 7), the value of the quantum energy obtained in this case was 2*6 x 10® eV. This is again in agreement with the quantum energy 2-37 x 10® eV expected from process (6). I t has been shown therefore th at between 400 and 900 kV there is one resonance (peak intensity at 480 kV) for the process (6), and between 300 and 950 kV one resonance (peak intensity at 570 kV) for process (7). A further confirmation of the above interpretation has been made in the following manner. The excitation function of the emitted y-radiation was measured with 6*2 mm. of aluminium between the counters (curve 3) fig. 6. The resonance a t 480 kV was thus reduced by a factor 30, whereas the resonance a t 570 kV was only reduced by a factor 4.
From the measured intensity of radioactive 18N the yield of y-ray quanta per incident proton in reaction (6) may be deduced. The total number of positrons emitted from a carbon target after bombardment with a known proton current for a known time interval was measured, correction being made for the absorption of electrons by the counter walls, decay of the radioactivity during bombardment, etc. The yield of positrons per incident proton was thus found to be 2-5 x 10-10. A recent measurement of this yield by Roberts and Heydenburg (1938) gave the value 3*5xlO_10» Assuming th a t for each proton captured a single quantum of energy 2*4 x 106 eV is emitted, the yield of quanta per incident proton is also, according to our measurements, 2-5 x 10-10. Some support for this assump tion may be derived from the half-value absorption thickness (curve 2, fig. 7) which indicates a quantum energy of 2-2 x 106 eV.
The yield of quanta per incident proton at 570 kV, in view of the fact th a t 13C is about 1 % abundant in natural carbon, appears from the excitation curve 1, fig. 6, to be about 100 times greater than the value 2*5 x 10-10, which relates to capture by 12C. Correction must be made, however, both for the fact th a t the proton capture by 13C results in the formation of excited 14N which often returns to the ground state by the emission of two quanta of energy 4 x 106 eV (Rose 1938), and for the different efficiencies of the counting system for the different energies of the y-rays associated with the two resonances. W ith these corrections the y-ray yield from process (7) is 6*2 x 10~9 quanta per incident proton. The ratio of the cross-sections for proton capture by 13C and by 12C is therefore approximately 25/1.

he excitation of y -r a d i a t i o ni n processes of proton capture 283
(e) Fluorine McMillan (1934) and Crane and Lauritsen (19346) first observed the y-rays emitted in the process of proton capture by fluorine. The excitation function of the capture radiation has been measured by Hafstad and Tuve (1935). Hafstad, Heydenberg and Tuve (1936) and Herb, Kerst and McKibben (1937) have confirmed some of the earlier observations. The existence of sharp resonances at 330, 890 and 940 kV was established. Between 400 and 700 kV the form of the excitation curve suggested the presence of several weak resonances, but no conclusive results were obtained owing to the difficulty of preparing a suitable target. We have studied the excitation function for the capture process for bombarding voltages between 300 and 1000 kV. The results obtained with very thin fluoride targets are shown in fig. 8. Curve 1 shows the TOOkV 900 kV results obtained in a run over the whole range of voltage, readings being taken in steps of 30 kV. I t may be seen th at confirmation of the earlier results has been obtained in the observation of resonance peaks at 340, 860 and 920 kV. The presence of two further peaks between 500 and 700 kV was also suggested. In view of the possible existence of other resonances in this region of voltage we have made a more detailed measurement of the counting in the background itself was 1500 coincidences/min. The initial intensity in this case could not be measured on any of the counting meters available. The end-points obtained from these two measurements were almost identical. I t is clear th a t there is no need to employ low rates of coincidence counting when sufficient intensity is available. I t may be A lum inium (mm.) A lu m in iu m (mm.)

Fig. 9
remarked th a t the time required in such cases to determine the y-ray energy limit is about 1 hr. W ith softer radiations, however, having quantum energies of about 106 to 2 x 106 eV the end-point determination is rendered more difficult by the fact th at many of the /?-rays which are produced pass less readily through both counters. For the conversion of the measured end-point absorption values into quantum energies a calibration curve ( fig. 10) was constructed in the following manner. The results of Bernardini and Franchetti (1937) were plotted (open circles). The end-point for the y-rays from radio-sodium, of quantum energy 3*08 x 10® eV was measured and found to lie upon the straight line passing through these points. The measured end-point value (28*0 mm.) for the capture radiation from boron under bombardment with protons of energy 850,000 eV was taken to correspond with the value 16*7 x 10® eV, which was found by Fowler, Gaerttner and Lauritsen (1938) for the quantum energy of the y-radiation. I t may be seen th a t this point lies upon the linear extrapolation of the points relating to the lower energies. We have also plotted upon this diagram other points determined by our measurement of end-points and values of quantum energies calculated from known mass values. This calibration curve seems therefore to be essentially correctfor quantum energies up to 17 x 10® eV. Theline does not pass through the origin, since the absorption of the counter walls was equivalent to about 0*6 mm. of aluminium.

S u m m a r y
Using a pair of Geiger-Mueller counters in coincidence the excitation functions for the y-radiations emitted in the process of proton capture by each of the elements, beryllium, boron, carbon and fluorine have been determined for energies of the incident protons from 0 to 10® eV. Several of these resonance peaks had not hitherto been discovered. In the cases of beryllium and boron a continuous excitation of y-radiation over the whole of the region of voltage employed was observed, and evidence is presented in support of the view th at this radiation must be associated with the capture process.
The yield of y-radiation was measured in the case of carbon, the values obtained being 2-5 x 10~10 quanta/proton for 12C and 6*2 x lO -9 quanta/proton for 13C.
A calibration curve is given relating the end-points of the coincidence absorption curves (obtained by placing aluminium between the counters) with the maximum quantum energies of the incident y-radiations.
A search for the production of radioactive 11C as a result of the capture of protons by boron has shown th a t the cross-section* for this effect is at least 100 times smaller than the value which has been reported by other workers.