Hess's Balloon Flights: How Cosmic Rays Were Found (1912)

In the summer of 1912, an Austrian physicist named Victor Hess climbed into the basket of a hydrogen balloon at Aussig on the Elbe with three electrometers and rose to about 5,350 meters (17,500 ft). Again and again he read how fast the air around his instruments was conducting electricity. Near the ground the rate was falling with height, as everyone expected. Above 1,000 meters it turned around and climbed, and at 5 kilometers it was more than double the ground value. Hess concluded that radiation of very high penetrating power enters the atmosphere from above, and that discovery earned him the 1936 Nobel Prize in Physics.

This article is about the experiment itself: the instruments, the seven flights, the numbers, the reasoning that ruled out the ground and the Sun, and how long it took the rest of physics to agree. For what cosmic rays are made of and where they come from, see our explainer on cosmic rays and particle physics.

A leak nobody could plug

The story starts with a nuisance. A gold-leaf electroscope is a simple device: two thin leaves of metal, or in later designs fine fibers, hang from a charged rod inside a sealed case. Like charges repel, so the leaves stand apart. If the air inside conducts even slightly, the charge leaks away and the leaves slowly fall together. Physicists around 1900 discovered that this happened no matter how carefully they insulated their instruments. Air, it seemed, was always being ionized, meaning that something was knocking electrons off gas molecules and leaving charged fragments (ions) that could carry the electroscope's charge away.

Radioactivity offered an obvious culprit. Henri Becquerel had found in 1896 that uranium salts emit penetrating rays, and Marie and Pierre Curie had shown that other elements did the same (our pieces on the Becquerel discovery and on Marie Curie's work on radioactivity cover that background). Uranium, thorium, and radium were known to be scattered through ordinary rock and soil. If the Earth's crust gave off gamma rays, the air near the ground would be ionized, and the electroscope would leak.

That explanation made a clean prediction. Gamma rays are absorbed by air, so the ionization should fall off as you climb away from the soil. Hess himself, working from his own absorption measurements and calculations by other physicists, estimated that ground gamma radiation accounts for only about three ions per cubic centimeter per second in his instruments, and that air absorbs most of it within roughly 1,000 meters. His instruments typically recorded around 16 ions per cubic centimeter per second at ground level, so the crust was supposed to explain a modest slice of the leak and then fade with altitude. The remainder was less clear, and that is where the experiment came in.

The central expectation: if radioactive rock is the only source, ionization should drop toward zero a kilometer or so above the ground. Hess found it did something else entirely.

Wulf, the Eiffel Tower, and the first doubts

Theodor Wulf, a German Jesuit priest and physicist, built the instrument that made the later measurements possible. In 1909 he replaced the gold leaves of the standard electroscope with two metal-coated glass fibers, a design sensitive enough to register ion production as low as about one ion pair per second, according to the historical account in Physics Today. Fibers made the instrument more compact and more rugged than a leaf design, which mattered once people began carrying them up towers and into balloons.

On an Easter visit to Paris in 1910, Wulf carried his electroscope to the top of the Eiffel Tower, about 300 meters (984 ft) above the ground. If the Earth were the source, the tower's summit should have shown a large drop in ionization. Wulf found only a modest decrease. In the English translation of Hess's 1912 paper, Hess cites Wulf's result as a fall that did not even reach half of the ground value at nearly 300 meters, far less than a purely terrestrial source predicts.

That was a hint, not a proof. A drop smaller than expected could mean many things, including that the radioactive material was more widespread than assumed, or that the radioactive gas radon and its decay products in the air were contributing. Radon is itself a product of uranium decay, so it seeps out of soil and drifts through the lower atmosphere. It was entirely possible that the tower result reflected the air, not the sky.

Gockel and Pacini: two more awkward results

Albert Gockel, a physicist working in Switzerland, took Wulf-type electroscopes aloft in balloons between 1909 and 1911. In December 1909 he made three flights in the balloon Gotthard, rising to 4,500 meters above sea level. He saw ionization decrease with altitude, but by less than a terrestrial origin predicted, and pressure changes inside his instrument made the readings hard to interpret (per the same Physics Today account). Hess's translated paper reports that Gockel concluded a non-negligible part of the penetrating radiation is independent of the direct action of the ground. Gockel's instrument problems left his results ambiguous.

The Italian physicist Domenico Pacini approached from the opposite direction. In 1910 he compared ionization on land and at sea near Livorno and in the Gulf of Genoa. In the work published in 1911 he sank an electroscope into the sea near the Italian Naval Academy in the Bay of Livorno. The discharge was significantly slower under water than at the surface. Water absorbs gamma rays, and the seabed is far enough away that crustal radiation should not have reached the instrument, so the leftover radiation had to be coming through the water from somewhere else. A translation and commentary by Michela De Maria and Alessandro De Angelis on arXiv notes that Hess cited Pacini's work in his 1912 paper.

By 1911 the situation was unsatisfying. Wulf and Pacini had signs that something other than the ground was contributing. Gockel's balloon data were ambiguous. And nobody had taken well-calibrated, ruggedized instruments high enough to settle the matter.

What Hess carried, and why the details mattered

Hess made two balloon flights in 1911, reaching about 1,100 meters, and found no essential change in radiation compared with the ground, a result recounted in CERN Courier's history of the discovery. Those flights taught him what to fix. The 1912 campaign was built around better instruments and a far more careful routine.

He used three electrometers, all of the Wulf type with two silicon-glass fibers. The English translation of his paper on arXiv describes them in detail:

Instrument	Vessel	Volume	Purpose
No. 1	Sealed, 3 mm walls, zinc-plated inside	2,039 cm³	Responds mainly to gamma rays
No. 2	Sealed, 3 mm walls, zinc-plated inside	2,970 cm³	Responds mainly to gamma rays
No. 3	Open to ambient air, thin 0.188 mm zinc walls	16.7 L	Meant to register softer, beta-like rays

Hess's three detectors in the 1912 flights. The sealed vessels kept their air at constant pressure during ascent, which was the flaw in several earlier balloon instruments.

The sealing is the important detail. A vessel open to the atmosphere holds less air at altitude, so it produces fewer ions for the same radiation, and a naive reading of its leakage rate gives a misleading picture. The two sealed instruments avoided that problem. Hess also zinc-plated the inside walls of the sealed vessels to reduce radiation from the wall material itself. The third, open instrument was corrected to a reference pressure of 750 millimeters of mercury, and Hess admitted in the paper that the correction was uncertain. He also applied no temperature correction, another limitation he stated openly.

To turn the fiber's drift into a physical quantity, Hess measured how many volts per hour each instrument lost, divided by the instrument's electrical capacitance, and converted the result into ions produced per cubic centimeter per second, using a value for the elementary charge of 4.65 × 10⁻¹⁰ electrostatic units. A custom sliding lens in the eyepiece, added by the instrument makers Günther and Tegetmeyer, sharpened the reading of the fiber's position. The instruments also lost some charge through imperfect insulation (0.2 to 0.7 volts per hour for the sealed pair), and Hess subtracted it.

The ground checks

Every flight began with hours of control readings on the ground. This is easy to skip past, but it is what turned the instrument readings into evidence. Hess needed to know what each device read at rest before the balloon left, so that any change in flight could be attributed to altitude, not to a drifting instrument. After treatment, the sealed instruments read about 16 and 11 ions per cubic centimeter per second at the ground, and the open one about 19 to 20 (after the pressure correction).

Seven flights in four months

Hess's paper, translated on arXiv, gives details of seven flights, most of them from Vienna. Because the high flight is the one everyone remembers, it is worth listing all seven. The table below draws on the translated paper and the commentary attached to it.

No.	Date (1912)	Launch	Max altitude	Notes
1	April 17	Vienna	about 2,750 m	Partial solar eclipse, cloudy; descended as gas cooled
2	April 26–27	Vienna, 11 p.m.	2,100 m	Night flight, about 11 hours, cloudless; landed near Brünn
3	May 20	Vienna, 10:10 p.m.	1,200 m	Night flight, short of ballast; landed near Königgrätz
4	June 3	Vienna, 9:45 p.m.	1,900 m	Storm forced landing at 1:30 a.m.
5	June 19	Vienna, 5 p.m.	about 950 m above ground	Hess alone, one instrument
6	June 28	Vienna, 11:30 p.m.	about 360 m	Five hours at 280 to 360 m
7	August 7	Aussig on the Elbe, 6:12 a.m.	5,350 m	Hydrogen balloon Böhmen; landed 12:15 p.m. at Pieskow, near the Schwielochsee

Hess's own text does not date flight seven; the August 7 date comes from the commentary and from the historical accounts cited here. The second flight used a 1,600 cubic meter balloon filled with illuminating gas (the coal gas that lit city streets), and the high flight used the 1,680 cubic meter hydrogen balloon Böhmen. Captain W. Hoffory piloted both, and the meteorologist E. Wolf flew along on the Aussig flight.

The Vienna flights, mostly flown at night, were not there for show. They were designed to answer a specific objection: that the Sun, not space at large, might be responsible. By flying at night, Hess could check whether the ionization changed when the Sun was below the horizon.

The flight that cannot be left out: the eclipse

On April 17, 1912, the Moon covered much of the Sun over Vienna in a partial eclipse. Hess ascended during the eclipse, from about 11 a.m. to 1 p.m., reaching roughly 2,750 meters in cloudy weather. If the Sun's rays were the source of the penetrating radiation, covering the Sun should have reduced it. Instrument No. 2 read 10.7 ions before launch and about 14.4 at 1,700 to 2,100 meters in the early eclipse, then 15.1 at around 50 percent coverage. The radiation did not drop. The balloon had to descend before totality-level measurements could be made, so the test was cut short, a limitation Hess did not hide.

He drew a careful conclusion. Together with the lack of a clear day-night difference in the night flights, the eclipse reading made the Sun an unlikely source of the penetrating radiation, at least as a direct, straight-line emitter of gamma rays. In his words, as quoted in CERN Courier, one can hardly consider the Sun the origin if no reduction appears at night or at an eclipse.

The numbers: down first, then up

Hess averaged the readings from all seven flights into altitude bands. Reproduced from the translation, the table runs like this (values in ions per cubic centimeter per second; "n/a" marks a band where instrument 3 gave no reading). The open instrument 3 is shown here before the pressure reduction:

Height above ground	Instrument 1	Instrument 2	Instrument 3 (not reduced)
Ground	16.3	11.8	19.7
Up to 200 m	15.4	11.1	18.5
200 to 500 m	15.5	10.4	17.7
500 to 1,000 m	15.6	10.3	18.5
1,000 to 2,000 m	15.9	12.1	18.7
2,000 to 3,000 m	17.3	13.3	22.5
3,000 to 4,000 m	19.8	16.5	21.8
4,000 to 5,200 m	34.4	27.2	n/a

Average ionization by altitude band, from Hess's 1912 paper as translated by De Angelis and Arcaro. At 4,000 to 5,200 meters the sealed instruments read about 2.1 and 2.3 times their ground values.

Read down the first column and the shape is plain. A small decrease near the ground, up to about three ions in some flights, matches the expectation from the crust: the soil's gamma rays are being absorbed. The dip is shallow and it levels out near 1,000 meters. Then the trend reverses. At 3,000 to 4,000 meters the sealed instruments gained roughly four ions, and at 4,000 to 5,200 meters they gained 16 to 18.

On the highest flight itself, the readings at about 4,700 meters on the way up were 40.7 and 31.8 ions for the two sealed instruments, against about 16 and 12 at the ground. CERN Courier summarizes the highest readings as 22 to 24 ions above ground values. On the descent at around 4,400 meters the readings were lower (28.1 and 22.7), and when the balloon landed at Pieskow they were back to 15.0 and 11.6, close to the starting figures. Some accounts say ionization at 4,500 meters was about three times the ground level; Hess's own band average puts the figure at about two, and the difference comes from whether one quotes the single highest reading or the band mean. Both agree on the direction and on the order of magnitude.

Not every reading was clean. On the descent, instrument No. 1 gave a suspect value at 1,200 meters that Hess attributed to stiff, thick fibers, and the open instrument No. 3 stopped working at 10:45 a.m. when its ionization cylinder came loose. That is why the high-altitude column for instrument No. 3 is empty. A reader skeptical of a single-balloon result had reason to ask whether two instruments agreeing was enough. They agreed closely in shape on every flight, which is the point of carrying more than one.

Ruling out the ground, the air, and the Sun

A rising curve is a surprise, but the claim that matters is the interpretation. Hess went through the candidate sources one at a time.

The Earth's crust. Gamma rays from soil and rock, he showed, account for about three ions in his zinc vessels and are largely absorbed within 1,000 meters of air. That fit the small decrease near the ground. It could not explain ionization that kept rising above 3,000 meters, where the soil is effectively out of range.

The atmosphere itself. Radon decay products, including the isotope then called radium C, can ionize air, and their concentration varies with weather. But variable local weather cannot explain a steady rise with height up to five kilometers. And Hess had a direct check: flights two and six spent hours in a stable inversion layer, where the air barely mixes, and showed no increase.

Instrument artifacts. Hourly fluctuations appeared at the same moment in several instruments, including on calm nights without weather changes. That pointed to a cause outside the instruments rather than to quirks of any one of them.

The Sun. The eclipse flight and the night flights, as above, left the Sun an unlikely cause.

What remained was the conclusion he put in print. As rendered in the English translation, it reads: radiation of very high penetrating power enters from above into our atmosphere. The phrase is hedged in a way worth noticing. Hess did not say where the radiation came from, only that it was coming down through the air from above and was not solar in any direct sense. The name "cosmic rays" arrived later. In the German-speaking world, terms such as Höhenstrahlung (altitude radiation) and Ultrastrahlung were used for decades.

One more point of honesty from the paper: over open water, other researchers (Wright, Simpson, McLennan, and Wulf himself) had measured bigger drops in ionization with altitude than Hess saw over land. Hess suggested that secondary radiation excited at the Earth's surface by the incoming radiation, stronger over land than over water, might explain the difference. He recorded the puzzle instead of hiding it.

What nobody could say yet

It is easy to read modern knowledge back into 1912. Hess did not know that most primary cosmic rays are charged particles. He assumed, like most physicists, that penetrating radiation meant gamma rays, since gamma rays were the penetrating radiation then known from radioactivity. The instruments he carried could not distinguish a gamma ray from a fast charged particle or from the shower of secondary particles that a primary ray sets off when it hits atoms in the upper air.

The question of what the rays are took until the 1930s to open up. Arthur Compton organized expeditions around the world in 1932 to measure the latitude effect, the dependence of the cosmic-ray rate on position relative to Earth's magnetic field. He showed that the effect exists and is larger for lower-energy rays, which means the radiation is at least partly made of charged particles that the magnetic field bends (per the Physics Today account). Millikan accepted the latitude effect after airplane measurements in 1933.

That timeline matters for how the 1912 flights should be judged. Hess correctly established the direction of the source (above, not below) and its penetrating power. He did not establish the identity of the radiation, and he could not have with a charge-collecting instrument.

How the field reacted

Hess reported the result at a scientific meeting in Münster about a month after the Aussig flight, and the finding was not immediately accepted. Several things slowed it down.

More flights, higher. Werner Kolhörster made five balloon ascents before the First World War, and the last, in 1914, reached about 9,300 meters, where ionization was roughly six times the ground value (Physics Today). That confirmed Hess's trend and extended it much higher; some sources quote a different multiple, such as nine times. The war then interrupted the work.

Millikan's resistance. In April 1924, Robert Millikan told the American Physical Society that the whole of the penetrating radiation is of local origin. Lake experiments in California, reported in 1926 and run with Harvey Cameron, changed his mind. Two meters of water absorbed about as much as two kilometers of air, and the radiation had to be coming from outside the atmosphere. Millikan then put his own name on the subject by coining "cosmic rays," a label that stuck over the German alternatives. Our feature on the Millikan oil drop experiment covers the earlier, more famous work from the same physicist.

Who gets the credit. The sequence of Wulf, Gockel, Pacini, Hess, Kolhörster, and Millikan makes a poor fit for a single-discoverer story. Historians have written whole papers on how nationality and the interruption of the First World War shaped who was credited (one is on arXiv). The Nobel committee in 1936 awarded Hess half of the physics prize "for his discovery of cosmic radiation," and gave the other half to Carl Anderson for the discovery of the positron.

Year	Who	What
1909–1910	Wulf	Wulf electroscope; Eiffel Tower measurement shows smaller drop than expected
1909–1911	Gockel	Three balloon flights, up to about 4,500 m; ambiguous results
1910–1911	Pacini	Electroscope discharges more slowly under the sea than at the surface
1912	Hess	Seven flights; ionization rises above about 1,000 m; reaches 5,350 m
1914	Kolhörster	Reaches about 9,300 m; ionization about six times ground value
1924–1926	Millikan	First rejects, then accepts an extraterrestrial origin after the lake tests
1932	Compton, Anderson	Latitude effect; positron found in a cosmic-ray cloud chamber track
1936	Nobel Committee	Hess and Anderson share the Physics prize

From balloon baskets to the positron

The payoff of Hess's discovery arrived in the next generation of detectors. In 1932, Carl Anderson at Caltech photographed particle tracks in a cloud chamber placed in a magnetic field and saw a track that curved the wrong way for an electron and was far too light to be a proton: the positron, the first antiparticle ever observed, most likely produced in a cosmic-ray shower. The Wilson chamber, which makes the paths of charged particles visible as lines of tiny droplets, is described in our pieces on the Wilson cloud chamber and on Anderson's positron discovery. In his Nobel lecture, Hess himself noted this connection between cosmic-ray tracks in strong magnetic fields and the positron.

Cosmic rays supplied particles far more energetic than anything a laboratory could produce in those years, which is why the Anderson track mattered. Muons, produced when cosmic rays strike the upper atmosphere, still reach the ground in large numbers and remain a classic demonstration of relativity; our piece on muon decay and particle-physics theory explains why.

Cosmic rays also leave a fingerprint on Earth. Interactions of cosmic-ray particles in the upper atmosphere produce the radioactive carbon-14 that archaeologists use for dating, a connection spelled out in our explainer on radiocarbon dating. The ionization that Hess detected in a balloon basket is the same physical process, one that has been going on in the atmosphere for as long as there has been an atmosphere.

What the experiment teaches about method

Part of why Hess's flights remain a teaching case has less to do with cosmic rays than with how to establish a surprising result with imperfect tools.

First, he made the control measurement a ritual. Hours of ground readings preceded every ascent. Second, he carried redundant instruments with different designs and checked that they told the same story in shape, even when their absolute values differed. Third, he went after the alternatives one at a time, with a test designed for each: eclipse for the Sun, night flights for the same, inversion layers for atmospheric emanations, extra absorption data for the soil. Fourth, he reported the awkward details, from the cylinder that came loose to the correction he called uncertain.

The campaign was also physically risky. Open-basket balloon flights to five kilometers mean thin air and cold, and an observer who has to read a microscope eyepiece through it. The Wikipedia summary of his career says he made the flights "at significant risk to himself." Hess took the measurements by day and by night, over several months, because the question required it.

Finally, the experiment shows that a good measurement does not need a theory in hand. Hess did not know what he had found. He knew where it was coming from, and that was enough to open a field.

Hess's later life

Hess was born on June 24, 1883, at Waldstein Castle in Styria, then part of Austria-Hungary, and earned his doctorate at Vienna in 1906. From 1910 he worked as an assistant at the Institute for Radium Research in Vienna, the setting for the 1912 flights. He was dismissed in 1938 after Austria was annexed by Germany and moved to the United States. He joined Fordham University in New York, became a naturalized US citizen in 1944, retired in 1958, and died in Mount Vernon, New York, on December 17, 1964, at the age of 81.

Questions people ask

Q: Why did Hess use a balloon instead of climbing a mountain? A: A balloon let him reach 5,350 meters in a single morning and take readings through the whole column of air, from the ground up. Mountains offered less altitude, and any reading on a mountain is contaminated by radioactivity in the rock around the instrument. Wulf's Eiffel Tower measurement showed how ambiguous a fixed structure can be. In the air, the nearest solid material was thousands of meters of empty atmosphere below.

Q: How high did Hess actually go? A: Hess reported 5,350 meters on the seventh flight, from Aussig. His altitude-band table runs to 5,200 meters, and secondary accounts round the figure to 5,000 or 5,300 meters. Everyone agrees on the flight and on roughly where it peaked; the spread comes from rounding and from which altitude figure each author chose to quote.

Q: Did Hess prove cosmic rays come from the Sun? A: No. He argued the opposite. The radiation did not fall during the April 17, 1912 partial eclipse, and it did not show a clear day-night difference, so he judged the Sun an unlikely origin. His eclipse flight was cut short by cloud and the need to descend, so the argument was careful rather than final. Later work established that most cosmic rays come from far outside the solar system.

Q: Who named cosmic rays? A: Robert Millikan popularized the name "cosmic rays" in the mid-1920s, after his own measurements in high-altitude lakes convinced him that the radiation came from outside the atmosphere. Before that, German-language scientists used terms such as Höhenstrahlung and Ultrastrahlung. Histories disagree on exactly when and where the English name first appeared, so a precise year is best treated cautiously.

Q: Why did Hess share the Nobel Prize with Carl Anderson? A: The Nobel Committee honored two connected discoveries in 1936: cosmic radiation, for Hess, and the positron, for Anderson. Anderson found the positron in a cloud chamber photograph of a cosmic-ray track in 1932, so the particle-physics work depended on the radiation Hess had identified.

Q: Were Hess's cosmic rays gamma rays? A: Hess assumed so, since gamma rays were the penetrating radiation then known from radioactivity. Compton's world survey in 1932 showed a latitude effect, which means the radiation is at least partly made of charged particles bent by Earth's magnetic field. The details of what the particles are belong to later work and to our article on cosmic rays and particle physics.

Q: Could anyone repeat the experiment today? A: Modern detectors replace the electroscope, but the logic is the same, and balloon-borne instruments remain a standard way to sample cosmic rays above most of the atmosphere. Hess's measurement is one of the landmark results gathered in our complete guide to science experiments, and the wider context sits in the complete science guide.

What this means

The flights of 1912 changed physics because Hess asked a very plain question, whether ionization falls with height, and then built the instruments and the routine to answer it honestly. The answer was the opposite of what a century of work on radioactivity predicted. What looked like a leak in a laboratory device turned out to be a measurement of the sky.

It also shows how discoveries tend to get made: in pieces. Wulf built the tool, Gockel and Pacini produced the hints, Hess made the decisive flights, Kolhörster pushed them to nine kilometers, and Millikan, Compton, and Anderson turned a curiosity into a research field. No one of them held the whole answer.

Sources

• Hess, V. F. (1912), "On the Observations of the Penetrating Radiation during Seven Balloon Flights," translated and commented by A. De Angelis and C. Arcaro: arXiv:1808.02927 (full text also at ar5iv).
• CERN Courier, "A discovery of cosmic proportions": cerncourier.com.
• APS News (April 2010) on Hess's balloon flight: aps.org.
• Physics Today, "A century of cosmic rays": physicstoday.aip.org.
• Pacini, D., "Penetrating Radiation on the Sea," translated by M. De Maria and A. De Angelis: arXiv:1101.3015.
• Wikipedia, "Victor Francis Hess," used for biographical dates: en.wikipedia.org.
• "Nationalism and internationalism in science: the case of the discovery of cosmic rays": arXiv:1012.5068.