Search

We describe the results of the GALLEX solar neutrino experiment after completion of the project. In particular, we summarize the results for GALLEX IV (12 solar runs) and for our 71As - spiking experiments. The integral GALLEX result for all 65 solar runs (GALLEX I–IV) is (78 ± 8) SNU (1σ). This is only slightly more than half of the expected rate. The significance of this deficit is assured. • directly by the results from our two 5Cr neutrino source experiments (at the 10% - level), and • indirectly by means of experiments in which 71Ge is generated within the target solution from beta-decaying 71As (at the percent level). GALLEX at Gran Sasso is now succeeded by Gallium Neutrino Observatory). The GNO status as of October 1998 is reported.

After 5.5 years of data taking, GALLEX ended its experimental phase. The solar neutrino production rate, 76.4 ± 8 SNU, represents about 55% of the predicted rate. The As tests prove, at the 1% level, the reliability of the technique and the detection of 51Cr neutrino, with the nominal efficiency (93.0±8%), control the response of the detector to neutrinos in the solar energy range.

Since May 1991, the GALLEX experiment — installed in the Gran Sasso Underground Laboratory — is continuously monitoring the solar neutrino signal by measuring the 71 Ge production rate in a target of 30.3 tons of nat Ga. The measured signal 77.1 ± 8.5 (stat) +4.4 – 5.4 SNU (1 σ ) [1] is well below the Standard Solar Model (SSM) prediction: ⋍ 120–130 SNU. To exclude unknown systematic effects as responsible for the observed deficit, an experiment with a (62.5 ± 0.4) PBq 51 Cr neutrino source has been performed 2. , 3. . The ratio R=0.97 ± 0.11 between measured and expected Cr-produced 71 Ge demonstrates that the deficit cannot be ascribed to unknown experimental systematics. Although astrophysical reasons could still remain as a possible explanation of the “solar neutrino puzzle”, we recall that the GALLEX result, when considered together with the results of the other solar neutrino experiments, would allow to strongly constrain the neutrino masses and mixing angles in the framework of the MSW effect.

The GALLEX collaboration aims at the detection of solar neutrinos in a radiochemical experiment employing 30 tons of Gallium in form of concentrated aqueous Gallium-chloride solution. The detector is primarily sensitive to the otherwise inaccessible pp-neutrinos. Details of the experiment have been repeatedly described before [1-7]. Here we report the present status of implementation in the Laboratori Nazionali del Gran Sasso (Italy). So far, 12.2 tons of Gallium are at hand. The present status of development allows to start the first full scale run at the time when 30 tons of Gallium become available. This date is expected to be January, 1990.

We report the GALLEX solar neutrino results for the measuring period GALLEX III, the period from 12 October 1994-4 October 1995. Counting for these runs was completed on 29 March 1996. The GALLEX III result (14 runs) is [53.9 ± 10.6(stat.) ± 3.1 (syst.)] SNU (1σ). This is 15.8 SNU below but statistically compatible with the new combined result for GALLEX (I+II+III) (53 runs), which is [69.7 ± 6.7(stat.) −4.5 +3.9 (syst.)] SNU (1σ) or (69.7 −8.1 +7.8 ) SNU with errors quadratically added. We also give the preliminary result from our second 51 Cr-source experiment: the measured detector response is 83±10% of expectation. The combined result from both GALLEX 51 Cr-source experiments is 92±8% of expectation.

The radiochemical GALLEX experiment, which has been measuring the solar neutrino flux since May 1991, has performed an investigation with an intense man-made 51 Cr neutrino source (61.9 ± 1.2 PBq). The source, produced via neutron irradiation of ≈ 36 kg of chromium enriched in 50 Cr, primarily emits 746 keV neutrinos. It was placed for a period of 3.5 months in the reentrant tube in the GALLEX tank, to expose the gallium chloride target to a known neutrino flux. This experiment provides the ratio, R , of the production rate of Cr-produced 71 Ge measured in these source exposures to the rate expected from the known source activity: R = 1.04 ± 0.12. This result not only constitutes the first observation of low-energy neutrinos from a terrestrial source, but also (a) provides an overall check of GALLEX, indicating that there are no significant experimental artifacts or unknown errors at the 10% level that are comparable to the 40% deficit in observed solar neutrino signal, and (b) directly demonstrates for the first time, using a man-made neutrino source, the validity of the basic principles of radiochemical methods used to detect rare events (at the level of 10 atoms or less). Because of the close similarity in neutrino energy spectra from 51 Cr and from the solar 7 Be branch, this source experiment also shows that the gallium detector is sensitive to 7 Be neutrinos with full efficiency.

The galliumchloride detector operated by the GALLEX-collaboration in the Gran Sasso Underground Laboratory responds primarily to pp-neutrinos. They are produced in the primary fusion reaction of hydrogen into deuterium and directly coupled to the solar luminosity. Standard Solar Models predict ca. 58% of the total signal expected in GALLEX (123-132 SNU) to be due to pp-neutrinos. The relative pp-neutrino dominance becomes even larger if the deficit of higher energy neutrinos (as observed in the Homestake- and Kamiokande experiments) is considered. During the first data taking period, 15 solar runs had been performed within the exposure period 14.5.1991 - 29.4.1992. The result, 81 {+-} 17 {+-} 9 SNU provided the first experimental evidence for pp-neutrinos from the Sun. At the same time, it confirmed the depression of higher energy neutrino fluxes relative to the model predictions. Here the authors report the results of 15 more solar neutrino runs, covering the period 19.8.92 - 13.10.93. They obtain 78 {+-} 13 {+-} 5 SNU. Evaluated together, the result for all 30 runs is 79 {+-} 10 {+-} 6 SNU. While the SNU rate of GALLEX I is well reproduced the statistical error has been reduced so substantially that a value of signal +2{sigma}more » is required to accommodate not only pp-and pep- but also the {sup 7}Be-neutrino induced {sup 71}Ge-production. Contrary, the fate of {sup 8}B-neutrinos has only little discernible effect on the GALLEX data. In conclusion, with the present errors GALLEX constitutes a 2.5 {sigma} problem for {sup 7}Be neutrinos within the frame of {open_quotes}astrophysical{close_quotes} solutions. Alternatively, the particle physics solution (MSW-effect) can consistently explain all available solar neutrino results, leading to a most probable mass scale with the muon-neutrino at approximately 3 meV (milli-eV). However, since the GALLEX result allows the presence of pp and pep neutrinos at full strength. the latter explanation of the data is not forced.« less

We report the solar neutrino results from the complete set of runs in the exposure period, GALLEX II, from 19 August 1992 - 23 June 1994. Counting for these runs was completed on 10 December 1994. The GALLEX II result (24 runs) is [75.2±9.7 (stat) −4.6+4.1 (syst)] SNU (1 σ). After three years of recording the solar neutrino flux with the GALLEX detector, the combined result from the 39 completed solar runs (GALLEX I+II) is [77.1±8.5 (stat) −5.4+4.4 (syst) SNU (1 σ) or 77.1 −10.1+9.6 SNU with errors combined in quadrature. The combined error (± 13%) has now approached a level where the limits on the derived contribution of 7Be neutrinos to the GALLEX signal confront the predictions of solar models.

The first period (GALLEX I) of data taking in the GALLEX solar neutrino experiment has been completed. From 14 May 1991 to 29 April 1992, the experiment observed the solar neutrino flux using neutrino capture by 71 Ga to form 71 Ge. Counting ended on 2 November 1992. The final result from the 15 runs of this periodpis (81±17 [ stat .]±9 [ syst .]) SNU (1 σ ). The first 6 runs of GALLEX II recorded the neutrino signal from 19 August 1992 to 3 February 1993. The preliminary result for this period is (97±23 [ stat .]±7 [ syst .]) SNU (1 σ ). With counting data considered until 29 April 1993, the joint result for all 21 runs is (87±14 [ stat .]±7 [ syst .]) SNU (1 σ ). The present neutrino recording period GALLEX II is continuing with one solar exposure every four weeks.

International audience; We have measured the rate of production of $^{71}$Ge from $^{71}$Ga by solar neutrinos. The target consists of 30.3 t of gallium in the form of 8.13 M aqueous gallium chloride solution (101 t), shielded by $\approx$ 3300 m water equivalent of standard rock in the Gran Sasso Underground Laboratory (Italy). In nearly one year of operation, 14 measurements of the production rate of $^{71}$Ge were carried out to give, after corrections for side reactions and other backgrounds, an average value of 83$\pm$19 (stat.)$\pm$8 (syst.) SNU (l$\sigma$) due to solar neutrinos. This conclusion constitutes the first observation of solar pp neutrinos. Our result is consistent with the presence of the full pp neutrino flux expected according to the "standard solar model" together with a reduced flux of $^8$B+$^7$Be neutrinos as observed in the Homestake and Kamiokande experiments. Astrophysical reasons remain as a possible explanation of the solar neutrino problem. On the other hand, if the result is to be interpreted in terms of the MSW effect, it would fix neutrino masses and mixing angles within a very restricted range.

The GALLEX collaboration is performing an experiment for the detection of neutrinos coming principally from the p-p fusion reaction in the Sun, via the reaction ve +71Ga →71Ge + e-. The experiment is running in the Gran Sasso Underground Laboratory of INFN, using as a target 30.3 tons of gallium in the form of 8.13 molar aqueous GaCl3 solution. A report is given of the status of GALLEX after the end of the operations devoted to the removal from the solution of the cosmogenically formed Ge isotopes, completed in the middle of 1991. The experiment is now collecting data on solar neutrinos and the data analysis is in progress. Preliminary results concerning the first year of measurement are presented.

The GALLEX result 83 ± 19 (stat.) ± 8 (syst.) SNU is two standard deviations below the predictions of stellar model calculations (124–132 SNU). To fit this result together with those of the chlorine and Kamiokande experiments requires severe stretching of solar models but does not rule out such a procedure, leaving the possibility of massless neutrinos. It clearly implies that the pp neutrinos have been detected. The Mikheyev-Smirnov-Wolfenstein (MSW) mechanism provides a good fit, and the GALLEX result fixes the Δm 2 and sin 2 2 θ parameters in two very confined ranges (around Δm 2 = 6 × 10 −6 eV 2 and sin 2 2 θ = 7 × 10 −3 and around Δm 2 = 8 × 10 −6 eV 2 and sin 2 2 θ = 0.6). Explanations of the solar neutrino problems based on the decay or magnetic interactions of neutrinos are disfavoured.

A radiochemical Gallium-Solar-Neutrino Detector has been set up at the Gran Sasso Underground Laboratory. All major installations are implemented and the target, 53.5 m3 of galliumchloride solution containing 30.3 tons of gallium, has been brought underground after removing cosmogenic Ge-isotopes through nitrogen purge. This final implementation status and the achieved parameters characterizing the experiment are described in this report. Full scale performance tests and optimalization of the operating conditions of the extraction system are presently under way.

A solar-neutrino experiment that uses /sup 7/1Ga as a detector can distinguish between broad classes of explanations for the discrepancy between prediction and observation in the /sup 3/7Cl experiment. A radiochemical experiment with the required amount of /sup 7/1Ga is feasible.

The formation cross-sections of 45Ca and 47Ca as a function of bombarding energy are computed. The nuclear reactions considered are the bombardment of 48Ti and 50Ti with neutrons and deuterons and the bombardment of 48Ca with neutrons, protons, deuterons, 3He and 4He particles. The calculations assume the formation of an excited compound nucleus with the subsequent de-excitation through nuclear evaporation. The effect of various assumptions is discussed. It is concluded that for the purpose of large scale production of 47Ca, the 48Ca(n, 2n)47Ca reaction is by far the most useful of all reactions considered.

The basic assumption underlying the analysis of compound nucleus reactions is the Bohr independence hypothesis which supposes the decay of the compound nucleus to be independent of its mode of formation. The applicability of the independence hypothesis to heavy ion reaction has previously been validated for the kinetic energy spectrum of the particles emitted from the compound nucleus 75Br formed by the reactions 16O + 59Co and 12C + 63Cu at one excitation energy(1). We have tested the validity of the independence hypothesis for heavy ion reactions by comparing the cross sections for the residual nuclei 28Al and 29Al from the reactions 18O + 12C and 16O + 14C over the excitation energy range of the compound nucleus 30Si of 30 – 40 MeV. We have also investigated the importance of the angular momentum in compound nucleus reactions of heavy ions by comparing the experimental formation cross sections of several residual nuclei from the above reactions and from the 18O + 18O and 18O + 14C reactions with statistical model calculations which neglect angular momentum effects and with calculations which take into account in a detailed fashion the effects of the angular momentum on the decay of the compound nucleus.

We have re-examined the decay of ${\mathrm{Rb}}^{83}$, which is reported to decay by electron capture, with emission of a single $\ensuremath{\gamma}$ ray of 525 keV, to the 1.86-h isomer of ${\mathrm{Kr}}^{83}$. With the aid of a lithium-drifted germanium detector and with a high-resolution beta spectrometer we find that the 525-keV photopeak observed with NaI crystals is really a superposition of three $\ensuremath{\gamma}$ rays at 521 keV (46%), 530 keV (31%), and 553 keV (16%). An additional $\ensuremath{\gamma}$ ray was found at 790 keV with intensity of 0.9%. Various coincidence experiments showed that only the 521- and 530-keV transitions lead to the 1.86-h ${\mathrm{Kr}}^{83m}$. The 553- and 790-keV transitions lead to the 9.3-keV level of ${\mathrm{Kr}}^{83}$. Direct electron capture to this level has also been obsered with an intensity of 6%. The fraction of ${\mathrm{Rb}}^{83}$ disintegrations bypassing the isomeric state of ${\mathrm{Kr}}^{83}$ is 23%. The following new levels were assigned to ${\mathrm{Kr}}^{83}$: 562, 571, and 799 keV. The total disintegration energy of ${\mathrm{Rb}}^{83}$ was shown to be \ensuremath{\ge}825 keV.

The formation cross sections of ${\mathrm{Li}}^{9}$, ${\mathrm{C}}^{16}$, and ${\mathrm{N}}^{17}$ from B, C, ${\mathrm{N}}^{14}$, ${\mathrm{N}}^{15}$, ${\mathrm{O}}^{16}$, ${\mathrm{O}}^{18}$, F, Ne, Na, Mg, and Al targets bombarded with 156-MeV protons were measured by means of their neutron emission. The results were compared with calculations based on the two-step spallation mechanism. Certain reactions were shown to provide direct comparison with prompt cascade calculation. In general, the agreement between theory and experiment was good, and it is suggested that some discrepancies are attributable to the neglect of the effect of nuclear structure in the calculations. The importance of $\ensuremath{\alpha}$ emission in determining the pattern of spallation product cross section was demonstrated.

The half-life of Kr 76 was determined by parent-daughter separation and by growth and decay of Br 76 . The value of t 1 2 = 14.8±0.1 h was obtained. Observation of the γ spectrum indicated peaks at ≈ 0.040, 0.073, 0.104, 0.135, 0.197, 0.267, 0.316, 0.360, 0.407 and 0.452 MeV. The most prominent γ rays are those at ≈0.040, 0.267 and 0.316 MeV. The rate of decay of the γ spectrum gave a half-life in reasonable agreement with that determined above. The Kr 76 nucleus seems to decay almost entirely by electron capture.

Excitation functions, angular distributions and differential ranges were measured for the 26 Mg( 18 O, 16 O) 28 Mg reaction at 18 O beam energies of 20–45 MeV. Excitation functions only were measured for the reactions 14 C( 18 O, 19 O) 13 C, 14 C( 18 O, 16 O) 16 C, 14 C( 18 O, 20 O) 12 C, 14 C( 18 O, 15 N) 17 N and 18 O( 18 O, 19 O) 17 O, 18 O( 18 O, 16 O) 20 O, 18 O( 18 O, 15 N) 21 F at 18 O beam energies of 13–41 MeV. We have identified these as direct reactions in which a single neutron, a two-neutron cluster, a deuteron and a triton are transferred between projectile and target. The cross sections for two-neutron transfer reactions were found to be relatively high and those for the 18 O+ 18 O and the 14 C+ 18 O reactions were higher than the ones of single-neutron transfers over most of the energy range. Attempts were made to apply the theory of Buttle and Goldfarb for single-neutron transfer to the case of two-neutron transfer in the 26 Mg( 18 O, 16 O) 28 Mg reaction below the Coulomb barrier. It is shown that for those reactions for which the assumptions, implicit in the model, are valid, good agreement is obtained with experiment. We also tried to apply the diffraction model of Dar and Kozlovsky to the calculation of the angular distribution of these reactions. A good fit to the experimental results could be obtained if quite different sets of parameters were used in the calculations for the two bombarding energies.

Experimental cross sections are reported for the formation of N/sup 13/ in the bombardment of Zn, In, Pb, and U with protons of 1.0, 1.9, and 2.9 Bev energy. These values are compared with theoretical N/sup 13/ emission cross sections for protons energies of 0.84 and 1.84 Bev. The calculations are based on the evaporation model. The previously described Monte Carlo procedure was modified in order to obtain better statistical accuracy for the calculated N/sup 13/ cross sections. Previously computed emission cross sections for He/sup 8/, Li/sup 8/, and Be/sup 7/ were also recomputed using the modified Monte Carlo procedure. The cross sections were computed for three different formulations of the interaction radius. Good fit with the experimental He/sup 6/ Li/sup 8-/ and Be/sup 7/ cross sections is obtained when the smaller values for the interaction radius are used. However, the fit with the experimental N/sup 13/ values is not good enough to exclude processes other than evaporation as contributing to the experimentally observed cross sections. (auth)

Monte Carlo calculations of nuclear reactions in the low-energy (E < 50 Mev) region are described. The calculations are based on the nuclear evaporation model of Weisskopf. Continuum theory was used for the calculation of inverse reaction cross sections. In the calculation of the level densities of excited nuclei, pairing and shell energy corrections were used in terms of charactertstic level displacements. The accurate equation rather than the approximate Maxwell distribution was used for the selection of the kinetic energy of the evaporated particle. Experimentally determined Q-values for the various reactions were used. The calculations are compared with experimental measurements for about 60 excitation functions of nuclear reactions in the mass range Cr/sup 50/-Se/sup 74/. Cameron's values for pairing energies were used at the outset; but a new set of pairing and shell energy correction values, which leads to substuntially improved agreement with the experimental curves, Is presented. The procedure which was used to arrive at this set is described and several features of the set are discussed. The need for a further downward correction of the level density of symmetrical (A = 2Z) nuclei is indicated. Computed excitation functions are shown for all the reactions studied as well asmore » for several reactions for which experimental data are not yet available. Further experiments on reaction cross sections are suggested which would allow a unique deterntination of the pairing and shell energy corrections of level densities for any value of Z and N in the region under discussion. The existence of a unique set of these correction terms would provide strong evidence for the validity of evaporation theory for the reactions considered. (auth)« less

The process of nuclear de-excitation by evaporation was calculated by using a Monte Carlo method and fast electronic computers. The competition between the emission of neutrons, protons, deuterons, tritons He/sup 3/, and He/ sup 4/ was studied as a functicn, of various parameters and experimental conditions. A systematic survey is presented for a range of atomic numbers, mass numbers and excitation energy of initial nuclei. The nuclides studied extend from A =49 to 239 with Z = Z/sub A/ plus or minus 3. The range of excitation energies for which computations are presented extends from 1OO to 700 Mev. Among the parameters affecting the process the importance of the temperature correction of the Coulomb barrier and the level density parameters are considered in detail. The average properties of the process such as average numbers of the various particles emitted, the average number of nucleons, of charged particles, and of charges emitted are summarized in graphical and tabular form for various combinations of the parameters used in the computation. Energy spectra of the various emitted particles are also shown. (auth)

Previous Monte Carlo calculations of nuclear evaporation reactions have been extended to include the emission of ${\mathrm{He}}^{6}$, ${\mathrm{Li}}^{6}$, ${\mathrm{Li}}^{7}$, ${\mathrm{Li}}^{8}$, and ${\mathrm{Be}}^{7}$ from Cu, Ag, Au, and Bi targets bombarded with high-energy protons (340-2000 Mev). Comparison with available experimental results shows good agreement in most cases. A discrepancy has been observed between the calculated and observed variation of ${\mathrm{Be}}^{7}$ formation cross section with the mass of the target nucleus, but even here the agreement is within a factor of three. It is shown that, for the usually chosen parameters of the calculation, a level density parameter of $a=\frac{A}{10}$ is necessary.

Delayed neutrons with a half-life of 11.4\ifmmode\pm\else\textpm\fi{}0.5 msec have been observed in the GeV-energy proton irradiation of targets of ${\mathrm{N}}^{15}$ and heavier nuclei but not from ${\mathrm{N}}^{14}$ and lighter nuclei. This new activity has been assigned to ${\mathrm{Be}}^{12}$. Cross sections for production of this nuclide from ${\mathrm{N}}^{15}$, ${\mathrm{O}}^{16}$, ${\mathrm{O}}^{18}$, F, Na, and Al have been measured. From cross-section systematics, it is estimated that delayed neutrons are emitted in about 7% of the decays. Upper limits for the delayed-neutron branches of ${\mathrm{He}}^{8}$ and ${\mathrm{B}}^{13}$ have been set at 3 and 0.3%, respectively.

Yttrium-83 was produced by the spallation of silver with 30-GeV protons and also by the 84Sr(p, 2n)83Y reaction with 39-MeV protons. The half-life of 83Y was determined by means of successive 83Sr daughter separations, and a value of 7·4 ± 0·1 min was obtained. In these experiments the half-life of 83Sr has been re-determined as 32·4 ± 0·2 hr.

The design of a Solar Central Receiver facility aimed specifically for a chemicals and fuel programme is described. The heliostat field, with a total reflecting area of 3500 m2 will deliver a maximum power of almost 3 MW(th). Several experimental stations are provided to give a flexible facility serving a variety of research projects. Some of these projects are outlined.

TRITYLMETHYL chloride, Ph3C.CH2.Cl, has interesting structural features, especially from the point of view of investigations of reaction mechanism1. We wish to record briefly the preparation, properties and reactions of this compound, and to refer to some related halides. The substance previously listed in the literature2 as tritylmethyl chloride is triphenylchloroethylene, Ph2C=C(Cl)Ph.

The radiochemical GALLEX experiment, which has been measuring the solar neutrino flux since May 1991, has performed an investigation with two intense 51Crneutrino sources (> 60 PBq) that were produced in the Siloe nuclear reactor and used at the Gran Sasso National Laboratory, one between June and October 1994, and the second between October 95 and February 96. The ratio, R, of the the neutrino source strength derived from the measured rate of 71Geproduction, divided by the directly determined source strength is R=1.01 +0.12−0.11 for the first source and R=0.84 +0.12−0.11 for the second one. The combined value of R for the two source experiments is R=0.93 ± 0.08. It shows that the > 40 % deficit of solar neutrino flux observed by GALLEX cannot be attributed to experimental artifacts and demonstrates the absence of any significant unexpected systematic errors at the 10 % level.

Previously, it was demonstrated that the GALLEX solar neutrino detector responds properly to low energy neutrinos, by exposing it to two intense 51 Cr-neutrino sources; the recovery yield of the product 71 Ge was reported to be 93%±8%. New experiments, in which known amounts of radioactive 71 As have decayed to 71 Ge in the full-scale gallium detector, strongly support this evidence. In several experiments, the gallium detector has been spiked with ∼10 5 71 As atoms, under varying conditions of how the 71 As was added (either carrier free, or with Ge carrier), how the gallium solution was mixed, and how long the 71 Ge remained in the gallium. 71 As decays by electron capture and positron emission to 71 Ge, with a half life of 2.72 d. Hot atoms are produced by these decay modes with kinematics that mimic solar neutrino capture, although the 51 Cr neutrino source provided a more perfect match. This relative disadvantage is offset by the much better statistics obtainable with the 71 As. In all 71 As experiments, the recovery of 71 Ge from the gallium was 100%, with uncertainties of only ±1%. The combined results from the 51 Cr sources and the 71 As spikes rule out any loss mechanisms for 71 Ge, including hot-atom chemical effects. Chemical processes in the aqueous gallium trichloride - hydrochloric acid solution guarantee that the 71 Ge atoms formed in the GALLEX target will be quickly converted to extractable, volatile GeCl 4 .

STERIC effects may be either thermodynamic or kinetic, that is, they may influence either reaction equilibria or reaction rates: this note is concerned with kinetic steric effects. Such effects depend on mechanism, and even the sign of a kinetic steric effect can be altered by a change of mechanism. ‘Steric hindrance’, obviously an inadequate title under which to summarize thermodynamic steric effects, thus becomes inadequate even to cover all kinetic steric effects.

Ina note under this title, Gerrard and Nechvatal1 have reported the conversion of neopentyl alcohol, (CH3)3C.CH2.OH (here written ROH), by thionyl and phosphorus halides into the esters RO.SOCl and RO.PCl2, inter alia. They remark that Whitmore and Rothrock2 demonstrated the resistance of neopentyl alcohol to the replacement of its hydroxyl group by halogen, through the agency of thionyl chloride and phosphorus tribromide. They add that we3 have attributed this to the steric hindrance of what they call “end-on approach” to the α-carbon atom. They state that Whitmore and Rothrock leave the impression that neopentyl alcohol does not react at all with the halides specified; but that the work of Gerrard enabled Gerrard and Nechvatal to predict the probability of an easy “broadside approach” of the reagents. We should like to add some points to this statement of the relation between the different investigations.

THE performance of a fractionating column recently constructed and used in this Department prompts us to place on record its chief characteristics. The column is packed with miniature (1/16 in.) -shaped ‘rings’ of phosphor-bronze gauze (100 mesh) and is enclosed in a vacuum jacket containing multiple, concentric aluminium reflectors. The length of the packed section is 12 ft. and its internal diameter is 20 mm. The column is supported on a specially designed ‘grid’. The ‘grid’ is a three-sided structure of metal piping built in the centre of a small laboratory, so that all points outside and inside the U-shaped assembly are easily accessible. The whole structure acts as a convenient supporting device, and the various metal pipes from which it is constructed serve to bring the usual services (gas, inlet and exit water) to numerous points suitably spaced for serving different types of fractionating columns and other items of equipment. The ‘grid’ also supports a large slate switchboard, from which all the electrical services can be controlled, and ‘hard’ and ‘soft’ vacuum lines are included in the assembly.