1 000 000 VOLTS
 Thanks to Paolo Brenni ( Author )  and to Willem Hackmann ( Bulletin's Editor)

Bulletin of the Scientific Instrument Society No. 63 (1999)

The Van de Graaff Generator


An Electrostatic Machine for the 20th Century

Paolo Brenni


A few months ago, the editor of the Bulletin asked me to give him my personal list of the '10 most important instruments' of the 20 th century. I asked for a few days of reflection. Some time passed but instead of having compiled a reasonable list of apparatus, my doubts and my hesitations had grown. Identifying the most significant scientific instruments up to the beginning of the 20 th century seemed clear enough, but ranking the apparatus invented since then was more troublesome.

Most 20th century scientific instruments are incomparably more complex than the ones of the previous centuries. A definition of them more precise than 'the hardware of science' (1) is almost impossible. Is a transistor or a magnetron a scientific instrument? They are, but they also are elements of more complicated scientific instruments. So in the list of the '10 best of the 20th century' should I include the magnetron or radar (whose electromagnetic waves are generated by the magnetron itself), or both ? And what about the computer? Is it a scientific instrument or not? Yes and no. Computers (also composed of thousands of instruments) are used for laboratory research and thus they can be well considered instruments, but they are also omnipresent in the domestic environment. In this last case is a computer a scientific instrument? If so, then must we also accept the hi-fi stereo and the kitchen microwave oven as scientific instruments? My presentation of the problem is perhaps paradoxical but certainly is not really far from the reality.

A Renaissance astrolabe, an 18th century Vacuum pump, or a sophisticated 19th century spectroscope, are self-contained and easily recognizable instruments. They have a well-defined archetype, which (like in the case of the astrolabe) could last for several centuries (2) . The miniaturization of electronic components created an enormous array of apparatus, which more or less looks all the same. Unlike the case of a 19th century spectroscopes, their exterior design is not distinctive anymore and it does not if all determine their use and their function. An inch square plastic chip is the sensor, which, periodically connected with a computer, substitute the classical recording thermometer, barometer or hygrometer. A black box can conceal an amplifier, a pulse generator, a photometer, a signal analyzer, etc. and so it is even difficult to understand from their look what they are.

Cataloguing these instruments will also create a series of problems. Because of their complexity it is unthinkable to describe them in the same detailed way we are used to with an ancient microscope or with a quadrant. Who made what? That is another very difficult question for the historian of 20th century science. If we admit that a signed Short telescope was made by the very famous English instrument maker of the 18th, century, and if we agree that a signed Ramsden sextant was at least manufactured in Ramsden's workshop (even if Ramsden himself perhaps never touched if), what will we be able to say about an electronic apparatus? Today the brand name on it does not mean much as far as production is concerned. We all know that many electronic instruments can be assembled in Korea, using microchips manufactured in the United States, electrical components coming from Spain, Taiwan or Germany, and finally sold under a British brand name. The problem with the apparatus of big science, which are often composed of tens of thousands of elements made by hundreds of different firms is even worse. Does it make sense to speak of a maker or manufacturer anymore? Probably less and less. Can we really speak of instruments or do we now have to speak of instrumental systems?

As you can see, my previous text is full of question marks, which correspond to open questions. Nevertheless, if my heart (and certainly those of many readers of the Bulletin ) is more attracted by the glitter of old fashioned 'brass and glass' than by the coolness of contemporary ,'aluminum and plexy', if is extremely important at the end of this century to seriously begin to think of its instrumental heritage. Finally, and I will return to this point in the closing remark of this article, a large part of this heritage risks being scrapped. But if we want to study, understand, and, when it is possible, to preserve the material witnesses of 20th century scientific enterprise, we must try to give some answers to the above mentioned questions. And that will probably be possible only by modifying in many cases our methodological approach to historical scientific instruments. I think it is an important and exciting challenge.

Finally, if I am not yet able to decide which are the 10 most important instruments of the 20th century, at least I have chosen to retrace the history of one of them: the Van de Graaff generator. Several reasons guided my choice. First, this apparatus strongly marked several decades of contemporary science and found application in various fields of physics, astrophysics, and medicine as well as in various industries. Second, it derived from a series of more ancient instruments. Third, if is useful and popular in schools and exhibitions as demonstration device. Fourth, depending on its use, size and power it can be considered a 'classical' instrument (as a demonstration electrostatic machine) as well as a typical artifact of 'big science' (3). Therefore the Van de Graaff generator can be considered a good example of emblematic 20th century apparatus.


The Van de Graaff electrostatic generator is an 'addition' electrostatic machine: charges are added to a conductor by a movable carrier. In its simplest and schematic form this apparatus is composed of a motor-driven vertical endless band (made of rubber, rubberized fabric, paper or another flexible insulating material) stressed between two rollers. The lower part of the band is electrically charged by a brush or comb, which is connected to a direct high-tension source. (In the smallest didactical generators the belt is simply charged by a small friction pad.) The charges, which can be negative or positive depending on the source, are carried by the pulley to a spherical hollow electrode which is installed at the top of the machine. Here, the charges are transferred from the belt to the sphere by a second brush or comb. Electric charges accumulate on the external surface of a conductor and thus the potential of the sphere is limited only by the corona effect and by the dielectric constant of the surrounding medium. The maximum electric tension, which in the first large air-insulated machines could reach a few million volts against the earth, is a function of the diameter of the spherical conductor.

The Van de Graaff generator, which was developed from the end of the 1920s and became immediately very popular, derives from a series 18th century electrostatic machines.

The Ancestors of the Van de Graaff Machine

The use of an endless-band of silk in the electrostatic generator can be found in the second half of 18th century (4). In 1784, Walckiers de St.Amand constructed such a machine with an horizontal looped silk strip passing over two wooden rollers.

The silk was rubbed by two sets of cushions fixed near the rollers. The prime conductor, where the charges accumulated, had two series of collecting points and was suspended centrally between the looped strip of silk. Walckiers also made a very large machine with a silk strip 1.5 meters wide and 7.6 meters long. This generator was successfully used at the Academie Royale of Paris and an improved version ( Fig.1) (5) of it was than constructed by the physicist Rouland (active in the 1770s and 1780s) who was the nephew and collaborator of the famous experimentalist Jean Rene Sigaud de Lafond (1740-1810).

Fig.1 Late 18 th century Rouland's endless-band electrostatic machine. From Gehler, op. cit. note 8

But this machine took up too much space and at the end of the 1780, the German Gottlieb Christian Bohnenberger (1732-1807) (6), the inventor of a well-known electroscope, proposed a new version of the machine with the belt set in a vertical position. In 1809 the French physicist Claude Veau Delaunay (1755-1826) illustrated in his physics treatise (7) the Walckiers' machine. Veau Delaunay admitted that this apparatus was very little used, because if was ,generally too big and not very nice looking ('..son aspect est peu agreable.'). Nevertheless he thought that it could be improved, and profitably used in various public institutions such as schools and hospitals, following the suggestions of the physician Louis Caullet de Vaumores (1743-?), who like Bohnenberger had proposed an endless-band vertical machine. In 1827 Walckiers' machine was again illustrated and described in Gehler's Physikalisches Worterbuch and even later in 1876 it was mentioned by E. Mascart as an historical curiosity (8). Unfortunately no endless-band machine of the time seems to survive.

In 1872 the young physicist Augusto Righi (1850-1920) (9) of Bologna in his PhD thesis described an 'induction electrometer'. (Fig.2) This apparatus not only used a looped flexible ring but was in fact a perfect miniature Van de Graaff generator ante litteram. However, this machine had not been conceived by his inventor as a generator but as a 'charge magnifier' for investigating weak electrostatic phenomena. With it, Righi wanted to investigate 'Volta's effect' and measure the weak potentials developed by the contact of different metals, hence the name 'electrometer'.

Fig.2 Righi's 'electrometer '. Collection of the Fondazione Scienza et Technica, Florence.

The idea of increasing an electric charge,which is too small to be measured, was not new. In 1786 Abraham Bennet (1750-1799) and in 1788 William Nicholson (1753-1815) had proposed their 'multipliers'. These apparatus, which in fact were induction electrostatic machines,represented fundamentally a mechanized version of Volta's perpetual electrophorus of 1775 (10).With these apparatus very small charges, too weak to be detected by a common electrometer, were 'multiplied' by electrostatic induction until they could be measured. Righi was working in the same direction when he proposed his machine, which is in fact an adder and not a multiplier (11).

Righi's apparatus is extremely simple (12). A rubber belt carrying a large number of brass rings rotates on two metallic pulleys. The lower one, which is insulated, is connected with a crank, the upper one is grounded with a copper strip. Close to the belt, in the neighbourhood of the upper pulley, there is a small metallic conductor (the inductor) which is connected to the weakly charged object to be studied. The inductor charges one after one another the brass rings of the belt which pass on the upper grounded pulley. Continuing their journey the rings enter a hollow insulated copper sphere, where they touch a third small metallic pulley fixed on its inside. Thus the charges of the rings accumulate on the external surface of the sphere. As the process continues the charges are continually added to the sphere. It is evident that this machine works in the same way as the Van de Graaff apparatus. Several Righi's apparatus were manufactured in Italy and can still be seen in various collections (13) but the idea was not developed for over 50 years. (Even if at the end of 19th century John Gray, an electro technician who wrote a famous treaty about electrostatic generator, proposed a rubber band induction machine which was more complicated than the Righi's ones). In 1893, Busch proposed an endless-belt machine. Between the looped belt of paper, rotating on two metallic cylinders, there was a unusual s-shaped toothed collector, which was connected to the prime conductor. But Bush's machine was simply a smaller and improved version of Rouland's generator, which was supposed to be a demonstration apparatus (14).Despite some minor modifications the endless-belt machines were never really popular and they could never compete with the disk induction generators of Holtz, Toepler, Voss, Carre,Wimshurst, Wommelsdorff and others.

The Race to High Voltages

In 1917 the British physicist Ernest Rutherford (1871-1937) transformed nitrogen atoms into oxygen by bombarding them with alpha particles generated by a radioactive isotope. The transmutation of elements, the mythical realm of alchemists, was at least on a microscopic scale, becoming a reality. But atom smashing requires very high energies. Natural radioactive elements such as the very expensive radium are sources of particles (alpha, electrons, as well as gamma rays) but their energy and their number are too low for penetrating the potential barrier (the Coulomb wall) of the nuclei of heavier elements. By the 1920s it appeared evident that further investigating of atomic and nuclear properties would require more energetic and intensive streams of accelerated particles. Charged particles could be obtained in different ways. Gas discharges could produce ions, while for electrons it was possible to use hot wire emission or other systems. The energy (E) of a particle in an electric field corresponds to the product of its charge (q) times the tension (U) of the field: E= q.U. Thus, a first possible solution of the problem was essentially to accelerate the particles in a vacuum tube to which a very high voltage was applied. The million-volt race had began, and it was rightly stated that: The high-tension accelerators stretched the power of insulators and the nerves of physicist to breaking point.(15) Several systems were proposed.

The Germans Brasch, Lange and Urban tried to use the atmospheric electricity of lightning, before turning to more practical and less hazardous impulse (or surge) generator. These kind of impulse generators were used by electrical engineers for testing electrical equipment. In 1930 a very powerful oil-insulated Tesla coil was built for the same purpose at the Carnegie Institution in Washington. In England, Cockcroft and Walton, who, in 1932, achieved the first successful disintegration of nuclei by electrically accelerated particles used a voltage multiplier with a complicated array of switches and condensers.(16) Resonance transformers, transformer rectifier and all manner of other apparatus were tested. Certainly, one of the best ideas was developed by Robert Jamison Van de Graaff, who chose to develop an old fashioned-style electrostatic machine. Finally, others (such as Lawrence with his cyclotrons) choose a completely different way: the particles could be accelerated in several steps using moderate electric fields. But that is another story.

The Invention and Evolution of the Van de Graaff Generator ( 17)

The American physicist Robert Jamison Van de Graaff was born in Alabama in 1901. After having received a masters degree in mechanical engineer in Alabama, he moved to Paris where he attended the courses of Marie Curie at La Sorbonne University (18). In 1925 he went to Oxford University, where, three years later he received a D.Phil. in physics. In 1929 Van de Graaff joined Princeton University as a National Research Fellow and at the end of the same year he built the first model (Fig.3) of his generator (80 kVolt).

Fig.3 Van de Graff demonstrating one of his early generator. Property of the Massachusetts Institute of Technology.

Soon the apparatus was improved and, in November 1931, he demonstrated for the first time a new, inexpensive and much more powerful machine (about 1-1,5 MVolt) at the inaugural dinner of the American Institute of Physics (19). Van de Graaff joined then the Massachusetts Technological Institute (MIT) as a research associate and in 1931 he began to construct a large double generator in an unused dirigible shed at Round Hill (South Dartmouth, Mass.) It consisted of two 23-foot high insulating columns each containing two belts and supporting an aluminium sphere, 6 feet in diameter ( Fig. 4 ).

Fig. 4 Drawing of the Van de Graaff generator at Round Hill (South Darmouth, Mass.) From Heilbron, Seidel, op. cit. note 4)

The columns were mounted on railway trucks so that the distance between them could be easily modified. This impressive machine, which was widely featured in technical journals and popular magazines of the time ( Fig. 5 ), was functional in November 1933. It was claimed that it could produce 7 million volts but in fact it developed about 5 MeV (20).

Fig. 5 Spectacular artificial lightning produced by the Van de Graaff generator at Round Hill. Property of the Massachusetts Institute of Technology.

Two small laboratories were located in the spheres, where scientists could study the effect in the accelerating tube which would have connected the two domes. Due to the difficulties of mounting the discharge tube between the spherical terminals this generator was never satisfactory as an accelerator. It was subsequently moved to MIT, where it was completely modified (21) and used for atom smashing and high-energy X-rays research. Finally in the 1950s it was donated to the Boston Museum of Science. In 1980, this generator, which is probably the largest surviving machine of this type, was installed in the Thomson Theatre of Electricity of the museum, were it is regularly demonstrated.

Before further retracing the development of the Van de Graaff generator, it is here worth mentioning another very large machine of this first type (double, non-pressurised). During the 1937 Paris Universal Exhibition, an impressive Van de Graaff ( Fig. 6) was installed in the newly opened Palais de la Decouverte, which was (and still is) located in the Grand Palais (22). This apparatus, built by A. Lazard under the direction of the famous French physicist Frederic Joliot (1900-1958), was supposed to be used after the exhibition as a powerful source of radioelements. This machine was composed of two Van de Graaff generators accumulating charges of different polarity at a total tension of 5 Mvolt. The Generators were 14 metres high and mounted on rails. The spheres at the top of them had a diameter of 3 metres. Each generator had three independent endless-belts driven by separate motors and charged by a 10000 volts direct current source.The system was entirely enclosed in a gigantic Faraday's cage.This machine, which amazed visitors to the fair with its spark several metres long was on the front-page of many magazines, but had unfortunately a sad fate. Because of World War Il it was forgotten in the Palais de la Decouverte and only in 1942 was it possible to undertake its removal to the Joliot's laboratory in Yvry near Paris. The machine had to be overhauled and a few mechanical pieces had to be substituted, but, due to the shortage and the critical wartime situation, nothing could be done. Therefore, this spectacular Van de Graaff was never used for any scientific research and it finally scrapped.

Fig. 6 Drawing of the Van de Graaff generator of the Palais de la Decouverte Paris , From Maury J.P., Le Palais de la Decouverte ( Paris , 1994) .

In 1935 Van de Graaff received a patent for his invention and he continued to work in the field of electrostatic generators. Together with his collaborator,John G. Trurnp, professor of electrical engineering at MIT, he was involved in the construction of such apparatus for producing highly penetrating X-rays for both medical and industrial purposes.

During World War II, Van de Graaff was director of the High Voltage Radiographic Project, where he developed electrostatic generators for the radiographic equipment of the U.S.Navy. After the war, Van de Graaff and Trump founded the High Voltage Engineering Corporation (HVEC), which became one of the most important manufacturers of electrostatic generators for cancer therapy, industrial radiography and research in high-energy physics.

In the late 1950s Van de Graaff invented the insulating core transformer which produced high voltage direct current exploiting magnetic flux instead of electrostatic charges. With his collaborators at the HVEC, he also successfully developed the technology of the tandem generator (see below). Van de Graaff remained associate professor of physics at MIT until 1960 and then he dedicated his activity to the HVEC. Besides several honorary degrees, in 1966 the American Physical Society awarded him the T.Bonner prize for his contributions to the development of electrostatic accelerators. Van de Graaff died in 1967. At that time over 500 particle accelerators of his type were in use.

In the 1930s the use of Van de Graaff generator spread rapidly, and their first, typical design changed very quickly. It is impossible to mention here all the technical findings, which contributed to increase the performances and the reliability of these machines. Studies on high voltage and on insulators, the realization of better materials and special mechanical elements certainly boosted the progress in the construction of electrostatic accelerators.

Several fundamental improvements and modifications were suggested by the American physicist and industrialist Raymond Herb (1908-1996) (23) . In 1931 Herb began his works with accelerators at the University of Wisconsin-Madison. In 1933, together with a few collaborators, he proposed one of the very first pressurised Van de Graaff generators (24).This modification proved highly successful and was further developed and universally adopted. All Van de Graaff machines for research and industrial purposes came to be enclosed in a (vertical or horizontal) cylindrical or cigar-shaped pressure tank, a departure from the typical column-and-sphere design. Electrostatic generators were becoming more and more sophisticated.

In the following years with his collaborators at the University of Wisconsin, Herb developed several pieces of equipment for the Van de Graaff generators. In 1935 he proposed the first column enclosed by closely spaced metal rings (equipotential rings), which contributed to produce evenly distributed voltage stress. In 1940 he and his collaborators introduced the use of three concentric high potential electrodes. During World War Il Herb joined the Radiation Laboratory where he worked on radar and after 1945 continued his work with the accelerators, developing ultrahigh vacuum techniques, negative ion beam formation, a corona triode controller, an electrostatic charging system and other devices. In 1965 he founded the National Electrostatics Corporation (NEC), which is still today one of the leading firms for the construction of Van de Graaff generators and ancillary equipment. Among the most interesting innovations proposed by Herb and his collaborator there is the Pelletron. This is fundamentally a Van de Graaff machine in which the rubberized-fabric endless belt is replaced by a special chain (or by more of them) of metal pellets connected by insulating nylon links (Fig. 10 ) .

Fig. 10 A section of a Pelletron charging chain . From the Pelletron web site, see note 25

In 1947 Herb and his group had put staples in a standard endless-belt to increase voltage stability. This idea was developed for several years and finally at the end of the 1950s and in the 1960s it was possible to find a good solution using a ' string of beads' charge carrier, which evolved into the modern Pelletron (Fig. 11 ) .

Fig. 11 Pelletron accelerator

The chain (which is curiously reminiscent of Righi's rubber ring with the brass carriers) has several advantages compared to the Van de Graaff belt, which is subject, for example to spark damages. The metal pellets of the chain are charged and discharged by induction (like many late 19th century generators) and no sparking and corona effects are involved in the process. Not only has it a longer life, but also it is also insensitive to moisture and gives excellent voltage stability. Today most powerful Pelletron chain generators can deliver current of 100-200 µA and the potential terminal can go up to 25-30 MVolt (25).A different type of charging chain called Laddertron, was developed by the HVEC. The name derives from the fact that the this chain had originally H shaped metallic carriers, which looked like a ladder.

Charge-changing, Tandem-principle and Tandem generators

But apart of a series of very important technical improvements, the discovery of the charge-changing effect, led to the realization of the so called tandem-generators, which largely improved the performances and increased the fields in which the Van de Graaff machine was used (26).The principle of this type of machine was independently proposed in the USA and in Germany.

Fig. 7 Bennet's early tandem accelerator. In the centre of the apparatus there is the high voltage cylindrical electrode with the foil stripper. From Rose, Wittkower, op. cit. note 25

In 1929 Bergen Davis (1869-1968) and Arthur Barnes at Columbia University had discovered that it was possible to electrically neutralise the positive charges alpha particles by attaching electrons to them. This fact opened the possibility of obtaining high energy by carrying these neutralized particles to a charged terminal and then, after having taken away the electrons, to accelerate them to ground voltage and to repeat the process again. The first experiments were unsuccessful but they attracted the attention of Van de Graaff. The discovery of negative ions (27) led Willard H. Bennet (1903-1987) in 1937 to propose a patent for an accelerator exploiting the charge - changing effect. In his apparatus ( Fig. 7 ) negative ions were accelerated to a thin metallic foil (positively charged), where the electrons were removed. The resultant positive ions were finally accelerated to a grounded target with an energy which corresponded to the one acquired in a 'classical' accelerator with twice the voltage. But at the time negative ions were rare, and Bennet's idea was soon forgotten.

It is certainly less well known that the principle of the tandem-accelerator was also discovered and developed in Germany by Harmut Kalmann (1896-1978), who had been an assistant and collaborator of the famous scientist Fritz Haber (1868-1934) (28). In the 1930 Kalmann, together with his collaborator Kuhn, were trying to improve a charge-change experimental accelerating vacuum tube (a modified 'Kanalstrahlen' tube) proposed by Christian Gerthsen (1894-1956). In 1930s Kalmann and Kuhn patented (29) an apparatus, which was quite similar to the one of Bennett. This device ( Fig. 8) produced positive ions, transformed them into negative ones, and accelerated them in to a positively charged tube electrode. The gas molecules flowing in the electrode changed by shocks the polarity of the ions, which, with a positive charge, were repelled by the electrode, thus being accelerated again. Kalmann and Kuhn tested their system but, unfortunately because of political reasons, which after 1933 forced the Kaiser Willelm Institute to a more utilitarian and war-oriented activity, the German discovery of the tandem principle did not produce any important application and was mainly forgotten by the history of science.

Fig. 8 The tandem accelerator of Kallman and kuhn. In the centre of the apparatus there is the high voltage cylindrical electrode and gas stripper. From: Weiss, op. cit. note 27.

A few years later, in 1951, the physicist Louis W. Alvarez (1911-1988) , who did not know Bennet's nor Kalmann's work, built a small charge-changing accelerator, showing the practical feasibility of such a system. Also in the 1950s, Van de Graaff and the HVEC finally developed a very successful machine of this type which became known as the tandem-accelerator or tandem-Van de Graaff. The first practical machine of this type was constructed by the HVEC for the Chalk River Laboratory of the Canadian Atomic Energy Agency.

Fig. 9 Scheme of a tandem Van de Graaff. Negative ions enter into the accelerator from the left side. From Rose, Wittkower, op. cit. note 25.

Tandem-accelerators machines are enclosed in high-pressurised tanks ( Fig. 9) . The voltage is generated by an endless-belt which is in a column (or horizontal tube) with equipotential rings whose open terminals are grounded while the cylindrical collector in the middle of the column is brought to a high voltage (we imagine +) by the charges carried by the belt. Negative ions produced by an appropriate source are accelerated in a vacuum tube inside the cylinder and in correspondence to the high voltage terminal the charge is changed by a gas or a metal foil 'stripper'. So, the now positive ions are repelled by the positive terminal and leave the accelerator with an energy which is double of that achieved with a single-stage Van de Graaff of the same voltage. This is an example of a well defined charge-changing (negative-positive) schemes but several others are possible (positive-negative; positive-neutral, neutral-negative, etc.) and multistage designs with two Van de Graaff in line one after another, are also used (30).The folded tandem design is constructed like a single stage vertical generator with one column containing two parallel accelerating tubes. One of them accelerates negative ions from ground potential to the high-terminal voltage terminal; while the other accelerates positive ions from the terminal down to the ground potential.

Tandem generators proved to be extremely efficient and reliable and they are utilized in several research and industrial laboratories around the world. Van de Graaff machines, especially the ones used in high energy physical research, can be very large. The most powerful ones, such as the French Vivitron (31) or certain types of tandem Pelletrons can reach a voltage of about 30 Mvolt. Vertical generators are usually installed in specially built towers. Furthermore, modern Van de Graaff generators are much more sophisticated than the ones built in the 1930s. Ion sources, particle injection and particle beam handling, vacuum (for the accelerating tube) and pressure equipment (for the tank), control and measurement instruments, etc., form a very complex array of apparatus typical of 'big science' equipment, for which the interchange of ideas between scientist and engineers proved to be particularly fruitful. In spite of the fact that other type of machines (synchrotrons, cyclotrons, etc.) can accelerate particles to much higher energy, electrostatic accelerators, which are very versatile and present great beam uniformity, lower cost, are ideal for many applications. The largest one are used for basic research (or as injectors for other kind of accelerators), while the smaller machines are employed in industry (neutron and X-rays production, ion implantation, polymerization, diagnostic measurements, etc.)

A Less Successful Type of Machine Based on the Van de Graff Generator

One of the most curious, through not really successful, modification of the Van de Graff generator had been conceived around 1936 in France by M.Pauthenier, professor of physics at La Sorbonne in Paris, and his collaborator, Mrs. Moreau Hanot ( 32) .In fact, they proposed to use a flow of charged dust particles circulating in a closed insulating pipe instead of the classical belt of the original apparatus. The dust was composed of glass spheres of a few microns in diameter. A blower produced a 60 in.p.s flow of these particles in the loop-pipe. For charging the dust there was a 'ionizer', which was composed by several wires parallel to the pipe. The wires were negatively charged (12,000 volts) by a kenotron rectifier and the electric fields of the wires ionized the gas molecules. The positive ions were immediately attracted by the wires, while negative ones were repelled in the direction of the walls of the pipe and charged the dust particles. Great care had to be taken for adjusting the various parameters (diameter and speed of the particles, voltage of the wire, diameter of the ionizing tube) so that the dust did not precipitate on the wall of the tube but could continue its journey after having been charged. At the top bend of the pipe the charged dust entered in a kind of centrifugal collector, which was connected with the spherical terminal electrode of the generator. In the collector the charges of the particle were transferred to the electrode.

At the Paris Universal Exhibition of 1937, together with the above mentioned large Van de Graaff, it was possible to see a Pauthenier-Moreau Hanot generator capable of producing a voltage of about 1.8 MVolt, working with a flow of minute glass spheres. Though this ingenious apparatus raised a certain enthusiasm, it was never widely used and remained a curiosity.

The Van de Graaff Machine as Didactic Apparatus

If the Van de Graaff generator continues to have a brilliant career in the field of high energy physics, astrophysics and in several industrial applications as research and professional apparatus, in its simplest form it also proves to be extremely popular as didactic instrument. Because of its solidity, of its simple construction, and of its insensitivity to air moisture, the Van de Graaff (in its basic and primitive design) became an ideal demonstration apparatus, overshadowing most of the old-fashioned and complicated electrostatic influence machine. Furthermore, for didactic purposes it is much simpler to explain the functioning of a Van de Graaff than one of the classical induction ('influence') machines. On the other hand, men have always been fascinated by lightning, big sparks, and effluvia of corona effects, therefore these machines always made for an ideal and very appreciated display it important exhibitions, science museums and science centers. Livingston contemplating the Van de Graaff in the Boston Museum of Science wrote:

The popular appeal of such gigantic generator has been tremendous. It is an awe-inspiring experience to stand beneath the huge spheres and feel the hair rise as potential increased, and then to see the long jagged strokes of man-made lightning as terminal discharges to the roof or down the column.(33)

So today every school laboratory as well as many museums use these generators for the most spectacular electrostatic demonstrations. In fact, a few didactic apparatus machines were developed recently on the principle of the Van de Graaff generator with some modification concerning the system of carrying the charges or transferring them to the belt. Most of these generators can reach a voltage in the range of 10-100 kVolt and are used for demonstration purposes.(34) Among them we can mention the ingenious generator conceived by the physicist Gabriel Lorente in Madrid.(35) Lorente's machine has four parallel rotating rollers with their surfaces kept in contact by means of springs. The two Internal rollers are respectively made of Teflon and nylon, while the external ones are made of metal. Teflon and nylon occupy opposite places in the 'electrostatic series' and due to their contact, the former receives electrons from latter. The charges on the surfaces of the insulating rollers are transferred to the metallic ones, which become respectively negatively and positively charged.

Final Remark

A final point has to be considered as far as large 20th century instruments are concerned. How many chances do they have to survive as material witnesses of the history of science and technology? Not many, I fear. The case of Van the Graaff generators is emblematic. As I mentioned above, at least one large historical Van de Graaff of the 1930s is preserved in a museum, where it attracts visitors with its spectacular high voltage phenomena. Small machines of this type used for didactic purposes are extremely common in the physics collections of educational institutions, and many of them will probably survive. But what about the large research generators built after World War Il ? In 1998, after 38 years of service at the Nuclear Astrophysics and Material Science communities, the Caltec EN Tandem Accelerator facility had been closed. The accelerator was destroyed : Not putting it too delicately: the machine was recently cut up and sold for scrap (36) .It is true that, machines of this type are extremely difficult to preserve. They are very large and heavy and they often occupy several hundred square meters laboratories. They are not particularly attractive: from the outside modern Van de Graaff generators look like big oil tanks. Often they cannot be dismantled without being destroyed. It is difficult and certainly far too expensive and complicated to keep them running just for didactic or historical purposes and, furthermore, they do not produce any particularly impressive phenomenon. In fact, modern Van de Graaff , in spite of the fact that they can generate tensions of several millions volts, are not supposed to produce sparks! Finally museums are not very keen to store this kind of equipment, which requires an enormous amount of space and which hardly can become an attractive exhibit. Of course if is much easier preserve documents such as photographs, films, videos, plans and drawings related to an apparatus, but the artifact itself has good chances to be scrapped. How much would we know about astrolabes if we had only written descriptions, engravings or images of them? Certainly much less than we know today.

Of course not all contemporary science is 'big science' and still many apparatus can be 'collectible' which do not create too many difficulties, but preservation of large equipment will present in the very near future a series of problems. On the other hand, if a large number of 'classical, ancient' instruments were rescued and preserved by private collectors, who played and still play today an essential role in the survival of artifacts, it is hard to imagine that they will be able to do the same as far as late 20th century big apparatus are concerned. Will the future instrument historians be forced to work without historical instruments as the most important sources of their research? It is a serious risk, which, I believe, we have to begin to consider at the dawn of the third Millennium.

Notes and References

1. This definition was given by John Burnett. See J. Burnett, The Hardware of Science, in 'Manual of Curatorship. A Guide to Museum Practice', ed. by Thompson J.M., Bassett D-A and others, Oxford, 1992 (II edition), pp. 374-391.

2. About scientific instrument archetype, Anna Van Helden recently presented a very interesting paper at the XIX Scientific Instrument Symposium (September 1999).

3. It is emblematic that a rock music band, which was quite famous a few year ago, was called 'Van der Graaff Generator'.

4. 1 just mention the machines that used a moving band, and I do not include here the fabric drum generators. See for a more detailed description W. Hackmann, Electricity from Glass (Alphen aan den Rijn, 1978), pp. 140-142 and the related bibliography.

5. Rouland, Description des machines electrostatiques a taffetas (Amsterdam, 1785).

6. He has not to be confused with Johann Gottlieb Bohnenberger (1765-1831), who proposed the gold leaf electrometer with two Zambini dry cells.

7. C. Veau Delaunay, Manuel d'electricite (Paris, 1809), pp. 14-16.

8. J.S.T. Gehler, Physikalisches  Worterbuch, III Band (Leipzig 1827), pp. 454-456 and E. Mascart, Traite d'electricite statique (Paris, 1876), vol. 1, p. 249. The machine is also illustrated in W. Nicholson, 'III. Bemerkung und Versuche, die Electricitat betreffend', Gilbert Annalen der Physik, 23 (1806), pp. 272-312.

9. Augusto Righi, became famous for his fundamental research on the properties of electromagnetic waves, which extended and refined the first observations made by Heinrich Hertz. Furthermore, Righi invented several apparatus. Among them a three-sparks oscillator, which was used in the very first wireless experiments of Marconi, who had followed some of the Righi's lessons at the University of Bologna.

10. See J. Gray, Les machines electriques a influence (Paris, 1892), pp. 161-163 and F. Luscia, La tecnologia delle macchine elettrostatiche (Brescia, 1928), pp. 88-98.

11. In an 'adder' the charges are accumulated in arithmetical progression, while in a 'multiplier' (such as the Thomson 'replenisher') the charges increase in geometrical progression.

12. A. Righi_'Descrizione di un elettrometro ad induzione', Il Nuovo Cimento, Serie II, VII, VIII, 1872, pp. 123-134; Luscia, pp. 184-186 and A. Righi, 'Sur le principe de Volta', Journal de physique theorique et appliquee, 3 (1874), pp. 1923. Righi also proposed a différent version of his machine.

13. See, for example, the collection of the Museo di Fisica of the University of Bologna or the collection of the Fondazione Scienza e Tecnica in Florence.

14. See H.W. Schmidt, 'Elektrisiermaschinen und Apparate', in Handbuch der Elektrizitat und Magnetismus, Vol. 1, L. Graetz, ed. (Leipzig, 1918), pp. 21-93.

15. See J.L. Heilbron, R.W. Seidel, Lawrence and his Laboratory. A History of Lawrence Berkeley Laboratory ,Vol. 1 (Berkeley, 1989), pp. 45-71.

16. Their apparatus was, in fact, not new. The Swiss physicist Heinrich Greinacher had proposed this kind of voltage multiplier in 1919.

17. The Van de Graaff generator was also mention as statitron in the past.

18. See Dictionary of Scientific Biographies; Mc Graw-Hill Modern Scientists and Engineers (New York, 1980), vol. 3, pp. 245-246; E.A. Burrill, 'Van de Graaff, the Man and His Accelerators', Physics Today, 1967, pp. 49-52; PH. Rose, 'In memoriam: Robert Jamison Van de Graaff', Nuclear Instruments and Methods, 60 (1968), pp. 1-3.

19. J. Van de Graaff, 'A 1,500,000 volt Electrostatic Generator', Physics Review, 38 (1931), pp. 1919-1920.

20. See for example the front page of Scientific American and the related article of Nikola Tesla (another pioneer of high-voltage technology) on this machine: 'Possibilities of Electro-Static Generators', Scientific American, 1934, pp. 132-134 and 163-165. Tesla stated : I believe that when new types [of Van de Graaff generators] are developed and sufficiently improved a great future will be assured to them. See also L.C. Van Atta., DL. Northrup, C.M. Van Atta, R.J. Van de Graaff., 'The Design, Operation, and performance of the Round Hill Electrostatic Generator', Physical Review, 1936, pp. 761-776. One of the best historical survey on the development of Van de Graaff generators and of others type of accelerators can be found in: S.M. Livingston, J.P. Blewett, Particle Accelerators (New York, 1962), pp. 30-72. The bibliography concerning electrostatic accelerator is so large that we can just suggest here some interesting contributions. Several interesting pictures of the large Van de Graaff can be seen at: http://www.mos.org/sln/toe/construction.html and http://www.mos.org/sln/toe/history.html. A very interesting site concerning Van de Graaff generators is proposed by Lyonel Baum, to whom I am grateful for much useful information.

21. The machine, which was enclosed in a metal dome was transformed into an uni-polar generator. The two spheres were joined together. In one of the columns were left the charging belts, while in the other the accelerating vacuum tube was installed.

22. A. Lazard, 'Le Palais de la Découverte scientifique. Le grand generateur electrostattique à 5 millions de volts', La Science et la Vie, 51 (1937), pp. 279-284.

23. See: Graw-Hill Modern Scientist and Engineers (New York, 1980), vol. II pp. 44-45 and G.A. Norton, J.A. Ferry, E.E. Daniel, G.M. Klody, 'A Retrospective of the Career of Ray Herb', in Heavy Ion Accelerator Technology: Eigth. international Conference (edited by K.W. Shepard) 1999, pp. 3-23. 1 am very gratefuI to Dr. G.A.Norton for providing me with much useful information about R.Herb and the NEC.

24. The breakdown voltage in insulating gas varies with the type of gas and its pressure. Several gases such as CCl4 and CCl2F2 were used in the electrostatic tank generator. Today the most used insulating gas is SF6, (sulphur hexafluoride).

25. See: http://www.pelletron.com and R.G. Herb, 'Pelletron Accelerators for Very High Voltage', Nuclear Instruments and Methods, 122 (1974), pp. 267-276.

26. See PH. Rose, A.B. Wittkower, 'Tandem Van de Graaff Accelerators', Scientific American, 223 (1970), pp. 24-33 and R. Van de Graaff, 'Tandem Electrostatic Accelerators', in Proccedings of the 1958 Accelerator Conference, Cambridge, Mass. (High Voltage Engineering Corporation, 1958).

27. Negative ions is produced by coupling an external electron to a neutral atom.

28. See B. Weiss, 'Hochste Spannung. Fritz Haber, Harmut Kallmarm and das "TandemPrinzip". Ein fruhes Kapitel des BeschleunigerGeschichte', Kultur & Technik, 1 (1997), pp. 4249.

29. German patent n° 696998, 9 February 1938.

30. It is impossible to mention here the complete bibliography concerning tandem generators. Therefore see the above mentioned references.

31. The Vivitron is a large (50-metre long tank) tandem Van de Graaff built in Strasbourg which was installed in the 1990s.[ Closed in 2003

32. See B. Kwal, M. Lesage, 'Les tres hautes tensions electriques par les poussieres chargees. Le generateur ionique de M. Pauthenier et de Mme. Moreau-Hanot', La Nature, 1937, 1 semestre, pp. 147-151.

33. See Livington op. cit. note 19, pages 32-33.

34. See for example the machine at the university of Nantes: http://www.sciences.univ-nantes.fr /physique/enseignement/tp /wimshurst.

35. I am very grateful to Dr. G.Lorente for the information he sent me concerning his machine. His generator was patented in 1991 in the USA (N. 4.990.813) and later in the European Union (N. 3665911). See http://info.uned.es/electrostatic-generator/ index.html.

36. For an account of the demolition of such a machine see http://www.its.caltech.edu/~arice/tandem.html.

Author's address

Fondazione Scienza e Tecnica - Via Giusti 2 - 50121 Florence - Italy

President of the Scientific Instrument Society (since july 2005)

© 1999 , Copyright Bulletin of the Scientific Instrument Society
Adaptation et actualisation pour l'Internet : Dr Lyonel Baum ( France ) © 2007

Version francaise ( French version )

Scientific Instrument Commission Home page

INDEX        |        VDG