The History of Volcanology

Haraldur Sigurdsson. Encyclopedia of Volcanoes. Academic Press, 2000.


This chapter is on the history of ideas about the generation of magma within the Earth—the root cause of all volcanic phenomena. It is a process that results in the transport of material and heat from depth to the surface of our planet. It would seem logical that the history of ideas on melting in the earth would be synonymous with the history of “classical” volcanology. This is not the case, however, because up to the mid-1970s volcanology was essentially a descriptive endeavor, devoted primarily to the geomorphology of volcanic landforms, geography of volcanic regions, and the chronology of eruptions. Until recently, volcanologists have shown little insight as to the physical processes of volcanic action and often ignored the causes of the melting processes that lead to the formation of magma. Volcanology thus became a descriptive field, lacking rigor, and on the fringes of science. It is now timely to redefine modern volcanology as the science that deals with the generation of magma, its transport, and the shallow-level or surface processes that result from its intrusion and eruption. Modern volcanology is, however, highly interdisciplinary and draws widely from diverse geoscience subspecialities.

The processes that bring about melting in the Earth had been discovered by physicists in the mid-19th century, but due to the bifurcation of scientific fields, volcanologists were on the whole ignorant of these findings of the early natural philosophers or geophysicists and continued to pursue a descriptive approach to surface features without acquiring an understanding of the fundamental processes at work deep in the Earth. Table I lists some of the principal historical advances made in the study of volcanism up to the 20th century.

The Beginnings

In prehistoric times, opportunities to “discover” fire and to observe heat abounded in the natural world. Lightning strikes ignited fires in the primeval forest or on arid, grassy plains. Volcanic eruptions and lava flows threw forth showers of glowing sparks and incandescent heat. Most likely these were the sources our early ancestors borrowed from, carefully nurturing the fires they kindled, for their sources were only fickle and local. With fire at their command, humans could venture into new realms of cold and dark—into polar terrain and mountain caves.

Our knowledge of human precursors and early humans is intimately linked with volcanic eruptions (see Chapter 80). It is because of the excellent preservation of fossils in volcanic deposits that paleoanthropologists have begun to unravel the mystery of human origins. Most of the oldest remains of early man come from volcanic regions in Africa and Indonesia, and the association of volcanic activity with these fossils is no mere coincidence. The most logical explanation for their relative abundance and good preservation in such environments is that the bones were rapidly covered by volcanic deposits. Just as the Roman cities of Pompeii and Herculaneum were instantly buried by the eruption of Vesuvius in 79 A.D. and preserved virtually intact to our day, so have earlier volcanic deposits sealed in the remains and fragile artifacts of our more distant ancestors, preserving them for millions of years. The stunning discovery of 3.7-million-year-old Australopithecus hominid footprints crisscrossing a volcanic ash deposit at Laetoli in the East African Rift is a graphic testament of the power of preservation by volcanic ash fall. Volcanic deposits also contain minerals that can conveniently be dated by measuring the decay of radioactive isotopes of potassium and argon, a method that has enabled the dating of fossil remains of the oldest human, Homo habilis, at Olduvai as 1.75 million years old. A similar setting in the African rift at Hadar in Ethiopia has yielded fossils of the possible human precursor, Australopithecus afarensis, permitting its dating at 3.5 million years. Volcanic stone is the earliest known material used in the creation of tools by genus Homo. The earliest stone artifacts are about 2.5 million years old, implements made from lava, found on the shore of Lake Turkana in east Africa. Following their humble beginnings as toolmakers with the fashioning of these crude “cores,” early humans gradually improved the techniques of working stone into flakes, choppers, handaxes, cleavers, and, finally, delicate obsidian stone blades in the upper Paleolithic.

More mundane materials such as pumice and volcanic ash have been used for nearly 3000 years in making cement. A mixture of volcanic ash and lime produces a very durable and water-resistant hydraulic cement. Volcanoes were also a principal source of sulfur. Early on people realized that sulfur burns with a strange and sputtering flame, giving off an evil-smelling and choking vapor. Probably the first uses for sulfur were to kill insects and to bleach wool, feathers, and fur. Homer knew of its fumigation property, and praised the “pest-averting sulfur” and Ulysses calledout: “Quickly, bring fire that I may burn sulfur, the cure of all ills.” Over the centuries the burning of a sulfur candle was a house-cleaning ritual following a case of contagious disease in the home. Sulfur features in Egyptian prescriptions of the 16th-century B.C., and Pliny the Elder describes its medicinal, industrial, and artisan uses in his Historia Naturalis (77 A.D.). The Romans also found a new application in warfare for the sulfur they mined in Sicily. By mixing it with tar, rosin, and bitumen, they produced the first incendiary weapon.

TABLE I The Evolution of Ideas about Volcanic Activity
Scientist Year Discovery or hypothesis
Anaxagoras 5th cent. B.C. Eruptions caused by great winds inside the Earth
Seneca 65 A.D. Combustion as a source of heat in Earth
Pliny the Elder 79 A.D. First scientific expedition to study an eruption
Pliny the Younger 79 A.D. First volcanic eyewitness account, Vesuvius eruption
Snorri Godi 1000 A.D. Basaltic rock in Iceland recognized as of lava flow origin
Liberti Fromondi 1627 Explosive eruptions compared to combustion of gunpowder
Edward Jorden 1632 “Fermentation” chemical reactions of sulfur generate terrestrial heat
René Descartes 1650 Incandescent Earth core as source of volcanic heat
Athanasius Kircher 1665 First global map of the distribution of volcanoes on Earth
Francesco d’Arezzo 1670 First experiments with molten rocks; melts Etna lava
J.-E. Guettard 1751 Identifies rocks at Volvic in Auvergne, France, as lava
Pierre Grignon 1761 Extracts crystals from glass slags; products of glass furnaces compared to products of volcanoes
Nicholas Desmarest 1763 Prismatic basalt columns in Clermont of volcanic origin
William Hamilton 1776 Pioneer in field volcanology; recognizes dikes and magmatic veins
Ben Franklin 1783 Volcanic eruptions have atmospheric effects and can cool Earth
James Hutton 1785 Recognizes intrusions of magma into the crust of the Earth
James Hall 1790 Melting and crystallization experiments on basalt
D. Dolomieu 1790 Some sediments are stratified and consolidated volcanic ash
Leopold von Buch 1799 Minerals in lavas formed by crystallization from magma
Lazzaro Spallanzani 1794 Earliest chemical analysis of volcanic rocks Water vapor is the dominant magmatic gas; causes explosions
John Playfair 1802 Speculates that change in pressure affects melting of rocks
Robert Kennedy 1805 First complete chemical analysis of basalts
Humphry Davy 1808 Exothermic reactions of alkalis provide volcanic heat
Pierre Cordier 1815 Identifies and analyzes the mineral components of lavas
A. von Humboldt 1822 Linear arrangement of volcanes related to deep-seated tectonics
George P. Scrope 1825 Decompression melting can produce magma in Earth’s interior Chemical differentiation produces variety of magma types
Charles Darwin 1835 Crystal settling is the agent of magmatic differentiation
S.-D. Poisson 1835 Recognizes that the high pressure in the deep Earth would lead to solidification of rock at higher temperature
Gustav Bischof 1837 Demonstrates experimentally that the transformation from solid rock to magma results in increase in volume
William Hopkins 1839 Decompression melting as a process in generation of magmas
Robert Bunsen 1851 Magma mixing produces intermediate magma types; recognizes that magmas are solutions
S. Waltershausen 1853 Water vapor expansion causes magma fragmentation and pyroclastic; recognizes importance of submarine volcanic rock formations
Robert Mallet 1858 Map of global distribution of earthquakes and volcanoes
Osmond Fisher 1881 Proposes convection currents within the Earth
John Judd 1881 Volcanic recycling of water from volcanoes to atmosphere, to oceans, and back to magma through solid Earth
Carl Barus 1893 First to determine the melting curve of basalt as a function of pressure
Alfred Lacroix 1902 Documentation of the eruption of Montagne Pelée
Frank Perret 1906 Pioneer in application of technology to study of volcanic phenomena
Alfred Wegener 1912 Theory of continental drift
Arthur Holmes 1915 Calculates a temperature profile for the Earth based on radioactive generation of heat
Arthur Holmes 1916 Basaltic magma generated by partial melting of periodotite
Arthur Holmes 1928 Decompression melting and magma generation by mantle convection
Norman L. Bowen 1928 Fractional crystallization theory developed to explain magma range

The Legends and Hell

Volcanic eruptions (Table II), the most spectacular and awe-inspiring of natural phenomena, have throughout history inspired religious worship and led to the creation of myths (Chapter 80). Even in the 20th century, these responses can be observed when a volcano erupts. James Hutton, regarded by many as the father of geology, was in 1788 the first to attempt to explode the myth associated with volcanic eruptions. In Hutton’s view,

[A]volcano is not made on purpose to frighten superstitious people into fits of piety and devotion; nor to overwhelm devoted cities with destruction; a volcano should be considered as a spiracle to the subterranean furnace, in order to prevent the unnecessary elevation of land, and fatal effects of earthquakes; and we may rest assured that they, in general, wisely answer the end of their intention, without being in themselves an end, for which nature has exerted such amazing power and excellent contrivance.

Primitive people the world over have long believed that volcanoes were inhabited by deities or demons that were highly temperamental, dangerous, and unpredictable. To appease the capricious gods, humans have for centuries made the ultimate sacrifice. Thus the Mayans, Aztecs, and Incas offered humans to their volcanoes. Nicaraguans long believed that their dangerous volcano Coseguina would stay quiet only if a child were thrown into the crater every 25 years. Similarly, young women were thrown into the crater of Masaya volcano in Nicaragua to appease the fire. Until recently, people on Java sacrificed humans on Bromo volcano, and still throw live chickens into the crater once a year. People living near the feared volcanoes of Nyamuragira and Nyaragongo in central Africa annually sacrificed 10 of their finest warriors to the cruel volcano god Nyudadagora. To those who were skeptical of such rites and pointed out that earlier sacrifices had failed to prevent or stop an eruption, the believers have countered with the argument that things would have been much worse without the sacrifice.

TABLE II Historic Eruptions
Eruption Significance
Thera (Santorini), 1650 B.C. One of the largest explosive eruptions on Earth; may have contributed to the decline of Minoan civilization
Vesuvius, A.D. 79 First major historical eruption to occur within range of major cities; first eruption to be documented by an eyewitness
Etna, 1669 Major lava flow event with widespread destruction in Catania from lavas
Lakagigar, 1783 Largest lava flow eruption on Earth; catastrophic impact on Iceland population from haze produced by volcanic gas emission
Asama, 1783 Local devastation from pyroclastic flows; country-wide impact in Japan from climate effects
Tambora, 1815 Largest historic volcanic eruption on Earth; highest eruption column (43 km); global sulfuric acid aerosol causes global climate change; largest known death toll of over 92,000 people
Krakatau, 1883 First large eruption of the modern age; severe impact and death toll from tidal waves (tsunami)
Mount Pelée, 1902 The classic example of a small eruption causing severe loss of life; the beginning of the study of pyroclastic flow and surge processes
Paricutin, 1943 The birth and growth of a new volcano observed
Bezymianny, 1956 One of the highest known eruption columns in a historic event (42 km)
Surtsey, 1963 Best evidence of magma-water interaction during explosive eruption
Mount St. Helens, 1980 First major eruption to be monitored intensively with modern technology
El Chichon, 1982 Most widespread pyroclastic surges from a historic explosive event
Nevado del Ruiz, 1985 Another example of a small eruption causing severe loss of life; tragic example of lack of hazard mitigation measures
Pinatubo, 1991 Largest eruption of the 20th century; major societal impact in Philippines; important global atmospheric effects

The Aztec people named the volcanoes surrounding the Valley of Mexico after their gods. Popocatepetl and Iztaccihuatl, lying to the east of the valley, were worshiped as deities, linked to a beautiful love story. When Popocatepetl (“Smoking Mountain”) returned victorious from war to claim his beloved, his enemies sent word ahead that he had been killed, and princess Iztacchiuatl (“Sleeping Woman”) died of grief. Popocatepetl then built two great mountains. On one he placed the body of Iztaccihuatl; on the other he stands eternally, holding her funeral torch.

Volcanic eruptions featured high in Greek mythology, which abounds with allusions to volcanoes, associating them with such gods as Pluto, Persephone, Vulcan, and the fearsome Typhon. The idea that volcanic activity represents the stirrings of the Titans, giants imprisoned in the Earth, goes back to the classical time of the Greeks. The mythical association of volcanic eruptions with battles between the Olympian gods and the Titans is very likely to date back to Preclassical antiquity. The Greeks saw the Titans as huge man-creatures, born of the Earth to attack the gods. Confined under various volcanic regions, their appearance on the land or in the air seemed the logical precursor to a volcanic eruption. The most awe-inspiring of all the Greek monsters was Typhon. The firstborn of Gaia (Mother Earth) and Zeus, he was the largest monster that ever lived. His arms, when spread out, reached a hundred leagues, fire flashed from his eyes, and flaming rocks hurtled from his mouth. Even the gods of Olympus fled in terror at his sight. Typhon rebelled against the gods and even opposed Zeus, who then hurled Mount Etna at him, trapping the frightful creature under the mountain. When Typhon was thus imprisoned under Etna, a hundred dragon’s heads sprang from his shoulders, with eyes that erupted flames, a black tongue and a terrible voice. Each time Typhon stirs or rolls over in his prison, Etna growls, and the Earth quakes, with eruptions and veils of smoke covering the sky.

The Hades of the Greeks and Romans made an easy transformation into the Hell of the early Christians, described in the Bible as a place with an eternal fire that shall never be quenched. St. Augustine spoke of Hell as containing a lake of fire and brimstone. By the Middle Ages, Hell had taken on great importance, with most scholars convinced that it was a real and fiery place. One place most often cited as the gateway to Hell was Mount Etna volcano in Sicily, and “sailing to Sicily” became a euphemism for going to Hell.

Wind in the Earth

The earliest known ideas on the cause of volcanic eruptions date to the Greek natural philosophers of the fifth-century B.C. Anaxagoras proposed that eruptions were caused by great winds stored inside the Earth. When these winds were forced through narrow passages or emerged from openings in the Earth’s crust, the friction between the compressed air and the surrounding rocks generated great heat, leading to melting of the rocks and the formation of magma. To anyone who has observed an explosive volcanic eruption, this is a perfectly logical idea, one that in fact was taken up by Aristotle and passed on by scholars until the Middle Ages.

Aristotle (384-322 B.C.) discussed the origin of earthquakes, attributing the same or similar origin for volcanic eruptions. “The Earth,” he wrote in the Meteorologica, possesses “its own internal fire,” which generates wind inside the Earth by acting on trapped air and moisture, leading to earthquakes and volcanic eruptions. He even makes a comparison with human anatomy in discussing the effect of the “internal wind”: “For we must suppose that the wind in the earth has effects similar to those of the wind in our bodies whose force when it is pent up inside us can cause tremors and throbbings.” Aristotle thought that the heat associated with volcanic eruptions was generated by friction produced when the wind moved rapidly through restrictions within the Earth. “The fire within the Earth can only be due to the air becoming inflamed by the shock, when it is violently separated into the minutest fragments. What takes place in the Lipari Isles affords an additional proof that the winds circulate underneath the Earth.”

The identification of volcanic activity with “wind and fire” by the early Greeks follows logically from actual observations of a volcanic eruption. The most striking phenomena are first the explosive uprush of hot gases or “wind” from the Earth through the crater; then the incandescent red-hot glow of molten rock, giving the appearance of fire. It was on the basis of these phenomena that Aristotle postulated a vast store of pent-up wind within the Earth, which generated friction and high heat and found its escape in volcanoes. After all, the Platonic view was that heat is a kind of motion. Aristotle’s concept of volcanism is thus firmly rooted in his ideas about the capacity of motion to generate heat as stated in his Meteorologica: “We see that motion can rarefy and inflame air, so that, for example, objects in motion are often found to melt.” This theory, which dates in fact to his predecessors, Anaxagoras, Plato, and Democritus, was remarkably long lived and had adherents well into the 16th century.

Internal Combustion

Initially the Roman philosophers adopted the Greek view of the causes of volcanic eruptions, but eventually they proposed another explanation, one more in keeping with the practical Roman mind. Lucius Annaeus Seneca (2 B.C.-A.D. 65) wrote on volcanism in his work on natural philosophy, Questiones Naturales. He attributed volcanic eruptions in part to the movement of winds within the Earth, struggling to break out to the surface, thus in part following Aristotle. But his most notable and truly original contribution to theories of volcanism is his proposal that the heat liberated from volcanoes is derived from the combustion of sulfur and other flammable substances within the Earth—an idea that was to have many adherents into and even beyond the Middle Ages. His claim was that there were great stores of sulfur and other combustible substances in cavities within the Earth, and that when the great subterranean wind rushed through these regions, frictional heating would set these fuels on fire. It is in his discussion of hot springs and thermal waters that Seneca first makes reference to sulfur and other combustible materials as a potential heat source, proposing that “water contracts heat by issuing from or passing through ground charged with sulfur.” He then extends this process to explain the blasts from a volcanic eruption in his work Questiones Naturales:

We must recognise, therefore, that from these subterranean clouds blasts of wind are raised in the dark, what time they have gathered strength sufficient to remove the obstacles presented by the earth, or can seize upon some open path for their exit, and from this cavernous retreat can escape toward the abodes of men. Now it is obvious that underground there are large quantities of sulfur and other substances no less inflammable. When the air in search of path of escape works its tortuous way through ground of this nature, it necessarily kindles fire by the mere friction. By and by, as the flames spread more widely, any sluggish air there may be is also rarefied and set in motion; a way of escape is sought with a great roaring of violence.

Seneca’s ideas were long lived and formed the basis of the interpretation of the causes of volcanic eruptions throughout the Middle Ages and even well into the 18th century.

Other significant writing on volcanism in the Roman world is found in poetry. The Latin poem Aetna, written between A.D 63 and 79, is of considerable importance in the history of volcanological thought; its likely author is Lucilius Junior. The poet maintained that the Earth is not solid, but has numerous caverns and passages. Heat, he argues, is more intense and powerful when in action in a confined space within the Earth, with the bellows-like action of winds in subterranean furnaces giving rise to volcanoes. In his view, the flames of Etna were fueled by a combustible substance, such as liquid sulfur, oily bitumen, or alum.

Philostratus the Elder (ca. A.D. 190) also discusses subterranean passages in the Earth, with fires breaking out through volcanoes: “If one wishes to speculate about such matters, the island of Sicily provides natural bitumen and sulfur, and when these are mixed by the sea, the island is fanned into flame by many winds, drawing from the sea that which sets the fuel aflame.”

The eruption of Vesuvius in 79 A.D. marks a watershed in the study of volcanism. When he observed the eruption from Misenum, at a distance of 30km, Pliny the Elder set out on the first expedition devoted to the study of a volcanic process; he died in the attempt. His nephew Pliny the Younger stayed at home, where he had a spectacular view of the eruption, and wrote the first eyewitness account of the phenomenon, a classic description in the volcanological literature.

Exothermic Chemical Reactions

Study of the Earth, like many scholarly activities, suffered a setback with the growth of the new Christian religion; and the only role of volcanoes in this new world order was to serve as a reminder of the hell fires burning below. This irrational attitude toward science continued well beyond the Middle Ages. Even by the 18th century, most writers on the philosophy of nature still considered that God created the Earth as a habitat for humans, that it had undergone certain stages of evolution since, and that it would ultimately be transformed or destroyed by Him. This was still manifest in the writings of John Wesley (1703-1791), the founder of Methodism, who taught that before sin entered the world, there were no earthquakes or volcanoes. These convulsions of the Earth were simply the “effect of that curse which was brought upon the earth by the original transgression.”

The Early Middle Ages of Western Europe coincided with a remarkable growth of learning in Asiatic countries, which were in their prime from A.D. 800 to 1100. Contrary to the Christian philosophy, the Koran encouraged the Islamic scholar to practice the mastery of taffakur, or the study of nature. Unlike most learned Europeans, who saw nature as a vivid illustration of the moral purposes of God, the Arabs sought knowledge that would give them power over nature. They were the first alchemists; and one element that figured prominently in their studies was sulfur, which was in part mined from volcanoes. Their discovery that sulfur could also give off heat in chemical reactions was to generate the idea that volcanic action was also fueled by sulfur in the Earth.

The early chemists dealt a death blow to the theory of internal combustion as the source of subterranean fires. In his work on Discourse of Naturall Bathes and Minerall Waters (1632), Edward Jorden (1632) rejected the notion that the Earth is a hollow and fiery furnace with a universal fire fueledby combustible coal, bitumen, or sulfur. He pointed out the fundamental problem regarding the internal fire: It would require a tremendous amount of air to keep it buring. Any flame that is confined without access of abundant air would soon be put out, because “fuliginous vapours … choake it if there were not vent for them into the ayre.” Such abundance of air could not reach the deeper portions of the Earth to fan the central fire. Instead, Jorden proposed that volcanic regions are underlain at only a shallow depth by “fermenting” material. As a source of the heat, he sought an explanation in the process of chemical reactions as a basis for a new system of the Earth. “Fermentation” or chemical reaction could take place in the presence of water, which was clearly abundant deep in the Earth’s crust, whereas combustion could not proceed in the presence of water. Although his reasoning logically should have ended the hypothesis of combustion as the heat source in the Earth, the idea persisted well into the late 18th century.

Another pioneer chemist was Robert Hooke (1635- 1703), who began his career as Robert Boyle’s laboratory assistant in Oxford. Hooke was one of the first to make a clear distinction between heat, fire, and flame, and, together with Boyle, to study the various phenomena of heat. He developed a “nitro-aerial” theory of combustion, in which thunder and lightning were likened to the flashing and explosion of gunpowder, during the combustion of sulfur and nitre. These ideas are reminiscent of the work of Liberti Fromondi in the Metteorologicorum (1627), in which the explosive power of earthquakes and volcanic eruptions was compared to the effect of gunpowder. Volcanic activity Hooke attributed to “the general congregation of sulfurous, subterraneous vapors.” In his work on Lectures and Discourses of Earthquakes and Subterraneous Eruptions (1668), Hooke also claimed that geologic activity was on the wane, and that subterranean fuel had been more plentiful in the past history of the Earth, when eruptions and earthquakes were apparently more severe: “That the subterraneous fuels do also waste and decay, is as evident from the extinction and ceasing of several vulcans that have heretofore raged.” This concept of a finite supply of “subterranean fuels” was widely accepted and was endorsed by Lord Kelvin in 1889.

Isaac Newton (1642-1727) concerned himself with the process of volcanism, but his ideas on the causes of this phenomenon were derived from his contemporary alchemists and early chemists. His ideas on the causes of “burning mountains” are clearly a direct outcome of his secret work on alchemy. In Newton’s chemical experiments, he had observed the evolution of heat, or exothermic reactions, when certain substances were mixed together, such as “when aqua fortis, or spirit of vitriol, poured upon filings of iron dissolves the filings with great heat and ebullition.” With some other sulfurous mixtures “the liquors grew so very hot on mixing as presently to sendup a burning flame,” and “the pulvis fulminans, composed of sulphur, niter, and salt of tartar, goes off with a more sudden and violent explosion than gunpowder.” By 1692 Newton had formulated opinions about the origin of heat in the Earth and its intensity, and compared it to the irradiance received by the Earth from the sun: “I consider that our earth is much more heated in its bowels below the upper crust by subterraneous fermentations of mineral bodies than by the sun.” In an experiment described in his work Opticks, Newton tested his fermentation hypothesis and drew his conclusion on the causes of volcanism:

And even the gross body of sulphur powdered, and with an equal weight of iron filings and a little water made into a paste, acts upon the iron, and in five or six hours grows too hot to be touched and emits a flame. And by these experiments compared with the great quantity of sulphur with which the earth abounds, and the warmth of the interior parts of the earth and hot springs and burning mountains, and with damps, mineral coruscations, earthquakes, hot suffocating exhalations, hurricanes, and spouts, we may learn that sulphureous steams abound in the bowels of the earth and ferment with minerals, and sometimes take fire with a sudden coruscation and explosion, and if pent up in subterraneous caverns burst the caverns with a great shaking of the earth as in springing of a mine.

A new and influential hypothesis on exothermic chemical reactions as a driving force of volcanism was developed with the discovery of the alkali metals by Humphry Davy. The new chemical theory did not call for a fluid interior, a vast storehouse of sulfur, or an internal fire in the Earth, and was consistent with the prevailing idea that volcanic activity in the Earth was progressively decreasing. Davy had a great interest in geology, eventually founding in 1807 the Geological Society of London. That year he was able to isolate by electrolysis for the first time the metals of the alkaline earths—potassium and sodium and later calcium, strontium, magnesium and barium. The high heat liberated upon the oxidation or “burning” of these metals upon reacting with water, with spectacular flames and explosions, so impressed Davy that it led him to propose a new hypothesis for volcanism in 1808. After his famous chemical discoveries, Davy was convinced that the heat given off during oxidation of the alkalies and the alkaline earths was the source of heat for volcanic action, supposing the existence of large quantities of the metallic alkaline earths within the Earth under volcanic regions.

One of the fiercest critics of Davy’s theory about volcanic heat was Gustav Bischof in Germany. The arrangement of volcanoes in great continent-wide lines on the Earth’s surface was taken by Bischof as a clear indication of a deep-seated origin of volcanic action, ruling out any superficial source within the crust. Davy had maintained that air could circulate through the Earth via volcanic craters and thus participate in heat-giving oxidation reactions. But the French chemist L. J. Gay-Lussac had shown by logic that it was impossible for atmospheric air to enter deep into volcanoes, because the pressure of the high-density magmatic liquid is acting outward. Air could not possibly flow into the volcanic system and fuel the oxidizing reactions against such a steep pressure gradient.

Charles Lyell in the Principles of Geology (1830) considered the possible role of oxidation of the alkalis and alkaline earths as a chemical heat source, following Davy’s ideas closely:

Instead of an original central heat, we may, perhaps, refer the heat of the interior to chemical changes constantly going on in the earth’s crust; for the general effect of chemical combination is, the evolution of heat and electricity, which, in their turn, become sources of new chemical changes. It has been suggested, that the metals of the earths and alkalis may exist in an unoxidized state in the subterranean regions, and that the occasional contact of water with these metals must produce intense heat. The hydrogen, evolved during the process of saturation, may, on coming afterwards in contact with the heated metallic oxides, reduce them again to metals; and this circle of action may be one of the principal means by which internal heat, and the stability of the volcanic energy, are preserved.

Even near the end of the 19th century the chemical theory was still alive.

Despite the many objections, the chemical theory of volcanism was surprisingly long lived and continued to be entertained by prominent scientists well into the middle of the 20th century. Arthur L. Day was the last of the proponents of the chemical theory, when he proposed in 1925 that chemical reaction between gases plays a leading part in generating volcanic heat. In his view, various gases from different sources in the Earth meet within the crust, and their reactions lead to local fusion and generation of magma. This idea was particularly appealing to Harold Jeffreys, a leading geophysicist in the early 20th century. He considered vulcanism as “local and occasional, not perpetual and world-wide.” In accounting for the “local and occasional” eruptions, Jeffreys appealed to Day’s chemical theory of heat generation and melting in the Earth.

The Cooling Star

During the latter part of the 17th century, several philosophers adopted the view that volcanism was due to original or primordial heat in the Earth. The first of these was the French mathematician René Descartes (1596-1650), whose works were to have a profound influence on thinking about the origin of the Earth. The solar system originated, he proposed, as a series of “vortices,” with the Earth first appearing as a star, “differing from the sun only in being smaller,” and collecting dense and dark matter by gravitational attraction and losing energy as matter falls into place by gaseous condensation. He divided the Earth into three regions: a core consisting of incandescent matter, like that of the Sun; a middle region of opaque solid material (formerly liquid but now cooled); and an outermost region, the solid crust. All of these layers had been arranged in this concentric fashion by virtue of their density. In this scheme, he maintained, there was enough primordial heat remaining to supply any volcano. Descartes’s ideas influenced Baron von Leibniz (1646-1716), a German mathematician who proposed that the Earth must have existed originally in a state of fusion, and had thus acquired its spherical form and concentric shell structure, with denser metals concentrated in the center. Lacking an independent source of heat, the planet had cooled by simple conduction over geologic time, forming a stony and irregular crust. The hypothesis of Descartes that the Earth contained a vast store of primordial internal heat was of fundamental significance in evolution of geophysics and volcanology.

Plutonists Victorious Over the Neptunists

Some rocks, which were later found to be of volcanic origin, were thought by the medieval philosophers to have originated by precipitation from water. Thus the Swiss philosopher Konrad Gesner (1516-1565) was convinced that the perfectly symmetrical and hexagonal basalt columns of many lava flows had precipitated as giant crystals from a primordial ocean. Two centuries later, Richard Pococke still claimed that the columns of the Giant’s Causeway, those huge six-sided pillars of basalt that rise from the sea in Northern Ireland, were formed by precipitation from an aqueous medium. The concept of a crystallization form for these highly ordered rock structures is not surprising, considering their regular shape: The pillars of the Giant’s Causeway resemble in many respects giant crystals that might have precipitated out of ocean waters. The debate on volcanic versus aqueous origin of basalt was to become one of the fiercest controversies in the history of Earth science.

The idea of precipitation of rocks from a great ocean had strong theological origins in the legend of the Great Deluge. The leading figure of the Neptunist school was Abraham Gottlob Werner (1749-1817). Among his students were the most famous geologists of his day, including Alexander von Humboldt, Leopold von Buch, Georges Cuvier, Johann Wolfgang Goethe, Jean François d’Aubuisson, and Robert Jameson. Werner considered volcanoes of minor importance, as accidental and relatively recent or postaqueous phenomena on the Earth. They owed their existence, he claimed, to the action of fire (much as the medieval philosophers had proposed), activated by the ignition of coal deposits or other flammable materials in the Earth: “Most if not all volcanoes arise from the combustion of underground seams of coal.” To support this notion, he proposed that deposits of coal or other combustible materials were invariably present in the vicinity of volcanoes.

By the first half of the 18th century, chemists were familiar with the formation of crystals in the laboratory as a product of precipitation from an aqueous solution. Because Werner and his followers had recognized that basalt was a crystalline rock, it was perhaps not unnatural that they also deduced its origin as a precipitate from an aqueous solution. It was not until after the middle of the 18th century that the idea slowly began to spread that crystallization could also occur as a result of the removal of heat from a silicate melt. In 1761 Pierre Clement Grignon complained that chemists overemphasized aqueous systems, pointing out, after he managed to extract crystals from glass slags, that the products of glass furnaces compared to the products of volcanoes. The idea that magma was a solution that could produce crystals was beginning to emerge, and a group of geologists embraced the view that basalt owes its origin to solidification of magma. The adherents of this theory became known as the Plutonists. A profound difference between the Neptunist and Plutonist theories related to the quantity and role of heat in the planet’s interior. The Neptunists saw a negligible role for heat as a geologic agent and considered volcanoes to be a minor phenomena related to shallow-level processes. The Plutonists, on the other hand, saw heat as the fundamental driving force, pointing out the abundance of volcanoes, basalts, and granite as evidence of melting of rocks at high temperature within the Earth. By the early 19th century, the Neptunist theory had become severely weakened and was encountering increasing opposition, especially when it was shown that the silicate minerals that compose crystalline rocks such as basalt and granite are insoluble in aqueous solutions at normal temperature.

The First Field Volcanologists

Most of the early ideas about heat in the Earth were based on “armchair geology,” speculation based on little or no field observation. In the 18th century, the approach to the study of the origin of the Earth gradually evolved from pure speculation to a search for answers in rock formations exposed at the surface and in mine shafts. It was the labors of certain men, including those connected with the great mining industry in southern Germany, that brought about the revolution in the study of the Earth and the invention of field geology. A great leap in volcanology was made when geologists were able to identify ancient rocks as of volcanic origin, in regions far removed from active volcanism. This breakthrough occurred in the middle of the 18th century in France.

In 1751 Jean-Etienne Guettard (1715-1786) and his friend de Malesherbes made a journey through the Auvergne region, where they noticed some very unusual, black, and porous stones in mile posts, and Guettard immediately suspected that they were lava rock. They then visited Volvic, where Guettard noted dipping layers of the rock with scoriaceous upper and lower surfaces and other unmistakable signs of volcanic activity. He concluded that the stones were indeed from a large lava flow. When they came to the area of Puy de Dome, Guettard recognized that basaltic rock of the hardened lava had flowed from an ancient crater in the cone-shaped hill above.

The geologist who first demonstrated the volcanic origin of basalt was Nicholas Desmarest (1725-1815). In 1763 he found prismatic basalt forming hexagonal columns in southwest Clermont. Tracing the rocks to their source, he found similar columns standing vertically in the cliff above, grading upward into the scoriaceous top of the lava flow. His simple observation of the association of columnar basalt as a component of a lava flow stands as a major advance in science. Although Desmarest had demonstrated unequivocally the volcanic origin of basalt, the debate between the Vulcanists and Neptunists continued to rage into the 19th century. His theory of basalt as lava was bitterly opposed by many.

Among the pioneers in field volcanology was William Hamilton, who resided in Naples for several decades and had ample opportunity to study Vesuvius. Among his many observations were precise drawings he made of the changing outline of the volcano during the eruption of 1767. Hamilton was also fortunate to be resident in Naples when the early discoveries and excavations were being made in Herculaneum and Pompeii, findings that brought to light the destructive power of the A.D. 79 Vesuvius eruption.

Experiments of the Plutonists

The undisputed leader of the Plutonists was the Scottish geologist James Hutton (1726-1797). He demonstrated the phenomenon of intrusion of magma into layered strata, and proposed that this molten rock originated in the highly heated interior of the deep Earth. Hutton devoted much energy to fieldwork, examining geologic strata in Scotland, England, and France.

Hutton was an intensely religious man and argued that the Earth’s processes followed a divine plan. Thus, volcanoes existed both as safety valves for the release of excess heat from within the Earth, and as a global force, because their expansive and uplifting power was needed to raise the surface of the Earth and contribute to the soils, so that plants and animals might flourish. Hutton thus saw volcanism as a means towards the realization of a higher purpose in the natural order of things.

Hutton and his followers had observed layers in the Earth’s crust that, while they resembled basalt in miner-a logical and other properties, were sandwiched between layers of sedimentary rock. They recognized that these rocks, which they referred to by the common mining term whinstone, were not volcanic but of igneous origin, and had been forcefully intruded between the sedimentary layers. Furthermore, they concluded that these rocks had solidified from magma.

Many of Hutton’s opponents had maintained that the melting and solidification of crystalline rocks, such as granite or gabbro, would only yield an amorphous, glassy mass upon cooling, as observed in glass-making furnaces, and therefore the process could not produce basalt, whinstone, or other volcanic rocks. This objection to the Huttonian theory was tackled experimentally by the Scottish geologist and chemist Sir James Hall (1761-1832). To test Hutton’s claim that rocks such as whinstone or basalt were derived from magma, Hall set out to do experiments on melting and cooling of basalts. First Hall melted the lava and basalt specimens, and then he converted the meltedrock to glass by quenching. Next Hall remelted the crystal-free glass and allowed the melts to cool slowly. In so doing he produced crystalline or partly crystalline rocks, rough and stony in texture, which resembled the original rock samples. These rocks, he showed, had melting and crystallization features exactly the same as those of basalts.

Demonstrating that melts of basaltic rocks precipitate silicate crystals upon cooling was of fundamental importance in establishing the volcanic theory on the origin of basalt. But if basalt rock was merely a product of solidification from a high-temperature silicate melt, then the experimental fusion and recrystallization of the rock should not affect its chemical composition. To test this, Hall submitted specimens of the products of his melting experiments to Robert Kennedy for chemical analysis in 1805, who found that their composition was essentially the same as the original rocks. Hall’s demonstration, which was of fundamental importance in establishing the volcanic theory for the origin of basalt, dealt a final blow to the Neptunist theory of an aqueous origin for basalt, granite, and other rocks of igneous origin. Although Hall was a pioneer in experimental work on high-temperature and high-pressure experimental petrology, he was not the first to conduct experiments with molten rocks. The first melting experiments on basalts were done by the Italian scholar Francesco d’Arezzo in 1670, who fused Etna’s lavas.

A Solid Earth

Most geologists studying volcanic activity in the early and mid-19th century continued to envisage a planet with a solid crust and a largely molten interior, influenced by the ideas of Descartes. Heat within the Earth was considered a residue of the central, primitive heat held by the Earth at the time of its formation, a supply that had been diminishing over geologic time. The idea of a central, primitive heat was developed primarily from the observation of increasing temperature with depth in the Earth, based on measurements in mines and drill holes. Such observations led to the theory that the Earth began as a molten sphere, and has been cooling ever since. But if the interior of the globe were molten, it should yield to the gravitational attraction of the Sun and the Moon, and there would be little or no relative movement of the Earth’s surface and the ocean’s, that is, no sea tide. The response of the Earth to tidal forces, then, became a critical test for the molten Earth hypothesis. The French physicist Andre-Marie Ampere (1833) was the first to point out that observations of tidal action did not support the hypothesis of a fluid interior.

From 1839 to 1842 William Hopkins analyzed the effects of the Moon and the Sun on the rotation axis of the Earth. He concluded that the outer rigid crust must be at least 1500 km in thickness: “We are necessarily led, therefore, to the conclusion that the fluid matter of actual volcanoes exists in subterranean reservoirs of limited extent, forming subterranean lakes, and not a subterranean ocean.” Hopkins’s studies were continued by his student, William Thomson (Lord Kelvin), who showed, on the basis of the effects on the Earth of gravitational attraction of the Sun and the Moon, that the Earth was essentially rigid and that the ideas of James Dwight Dana of “an undercrust fire-sea” were untenable. Geologists were now in a great quandry: There was no evidence for a large body of magma within the Earth, yet there was a requirement for supplying magma to countless volcanoes throughout the history of the planet. How can you derive magma from the interior of a solid planet?

The Earth’s structure at the turn of the century was regarded as a series of concentric, solid shells or layers, with an outermost 30- to 40-km-thick, low-density layer of granitic crust (sial) making up the continents. This was underlain by a very thick layer (sima or the mantle) of high-density, ultrabasic rock with the composition of perioditite, and finally by the central core. Those who were searching for a molten region in the Earth as a source of magmas were now faced with a geophysical picture of a largely solid interior. A magma source in the molten core seemed impossible, at a depth of more than 2900 km below the surface.

Decompression Melting

The discovery that the Earth beneath the crust was essentially solid completely eliminated the “undercrust fire-sea” that Dana and others had proposed as a source of the magmas that feed Earth’s volcanoes. Scholars, now faced with the task of showing how hot magma could be derived from within a solid Earth, found in the science of thermodynamics a partial solution to the problem. In the early years of the 18th century, it had been appreciated that pressure influences the temperature at which substances will undergo a change of phase—such as from a liquid to a gas, or from a solid to a liquid. The possible effects of pressure on melting of rocks in the Earth had been qualitatively appreciated very early on in the development of geologic thought. Perhaps the earliest recognition of this fundamental effect was made in 1802 by John Playfair. He pointed out that just as a change in pressure affects the boiling point of water, so would melting in the Earth be influenced by the great pressure exerted by the overlying rocks.

The next to address the importance of pressure in volcanic systems was George Poulett Scrope. By 1825 in his work Considerations on Volcanos, Scrope had realized the effect of pressure on the solubility of water in magmas, and was one of the first to point out that decrease of pressure on a water-rich magma could explain volcanic explosions due to the release of dissolved water. He also argued that a change of pressure could either lead to melting or crystallization of magma, without change in temperature:

Having thus far considered the effect of an increase of temperature, or a diminution of pressure, on a mass of lava under such circumstances, let us examine what will follow from the reverse; namely, an increase of pressure, or a diminution of temperature. Upon the solid lava, it is clear, no corresponding change will be produced; but every diminution of temperature, or increase of pressure, on a mass, or a part of the mass, liquified in the manner stated above, must occasion the condensation of a part of the vapour which produces its liquidity, and so far tend to effect its reconsolidation.

Simeon-Denis Poisson had proposedin 1835 that the excessively high pressure in the deep interior of the Earth would lead to solidification of rock material at much higher temperatures than at the low pressures near the surface.

While Playfair and Scrope were struggling with a qualitative approach to the question of the effect of pressure on melting, a more elegant and quantitative approach was being developed in France and Germany, marking the birth of thermodynamics. From the work of Sadi Carnot, Rudolph Clausius, and Benoit-Pierre-Emile Clapeyron came the Clausius-Clapeyron equation, which quantifies the relationship between temperature, volume and pressure of a substance:

Consider two phases of the same chemical substance, for example, a liquid and solid (magma androck), in equilibrium with one another at temperature T and pressure P. By supplying heat slowly to the system, one phase changes reversibly into another to bring about melting, with the system remaining at equilibrium. In the equation, the fraction dT/dP represents the rate of variation of the melting point T with change in pressure P. The value ΔV represents the volume change on melting and is generally positive (i.e., the specific volume of the solid is smaller than the volume of the corresponding liquid), while the value ΔH is the entropy change. The equation expresses the variation of pressure and temperature for a system in equilibrium. The magnitude and sign of the slope of the melting curve, dT/dP reflects the magnitude and sign of the volume change, ΔV, of the substance in question upon solidification or freezing.

It was shown early in the 18th century, in melting and crystallization experiments, that the specific volume of a volcanic rock such as basalt is lower than that of the corresponding magma. Many of the geologists who studied columnar basalt correctly interpreted its structure as evidence of contraction of magma upon cooling and solidifying to rock. During cooling and solidification, the material shrunk in volume, forming roughly hexagonal columns separated by a pattern of contraction joints. Thus it was clear that the term ΔV in the Clausius-Clapeyron equation was positive, and consequently, the equation predicts that the pressure-temperature melting curve (dT/dP) of the source rock of basaltic magma has a positive slope in the Earth, that is, the temperature of melting increases with pressure. This interpretation of columnar basalt was consistent with the experimental results of the German chemist Gustav Bischof in 1837, who carried out one of the first measurements on the volume change of basalt and other volcanic rocks upon fusion and showed that rocks contract upon solidification from magma. He concluded that a melt of granite rock contracted about 25% upon solidification, trachyte about 18%, and a basaltic melt 11%.

Another German chemist, Robert Wilhelm Eberhard Bunsen, was one of the first to experiment on the relation between pressure and melting point of substances. The laboratory facilities available in Bunsen’s time (1850) permitted only modest pressures to be achieved, roughly the equivalent of the pressure at a depth of 1 mile in the ocean. He therefore performed his experiments on materials with a relatively low melting point, such as ambergris and paraffin, and extrapolated these results to the high pressures and temperatures prevailing deep in the Earth. His work showed that a pressure increase of only 100 atm increased the melting point of these substances by several degrees centigrade. At about the same time (1851), William Hopkins began to experiment with James Joule, Lord Kelvin, and William Fair-bairn, on the effects of pressure on the solidification of the Earth’s interior. A large lever apparatus generated pressure up to 5400 atm, equivalent to the pressure at approximately 15-km depth in the Earth. Initially the experiments were on substances with low melting point, such as beeswax and spermaceti, and they essentially confirmed the work of Bunsen.

These pioneers had established that a solid hot substance such as rock at depth in the Earth could spontaneously begin to melt if the presure is decreased, without the addition of heat. They had finally found the secret of the generation of magmas in a solid Earth: decompression melting. But the solution of one problem led to the creation of another, even more daunting problem: What is the mechanism that brings about decompression? Some ingenious proposals were put forth to solve this problem. Thus in 1878 the American geologist Clarence King, realizing that local relief of pressure leads to melting in the Earth, suggested that the pressure release could occur with the erosion and removal of overlying crustal rocks: “So that the isolated lakes of fused matter which seem to be necessary to fulfill the known geological conditions may be the direct result of erosion.” This seemed perfectly logical at the time, but if true, the rate of erosion of a given area must occur at a higher rate than that of heat conduction from the rising hot rock. Geologic evidence did not support this theory on two counts. First, many mountainous regions, where erosion is highest, show no signs of volcanism; and second, the rate of pressure relief due to erosion is so slow that heat lost through conduction would prevent melting at depth.

As director of the U.S. Geological Survey, King fostered the fundamental experiments of Carl Barus on rocks at high pressure. Barus determined the volume change and latent heat of basalt upon melting. He concluded that the term dT/dp in the Clausius-Clapeyron equation was equivalent to 0.025 for basalt, that is, the melting point increases at a rate of 2.5°C for every 100-atm increase in pressure. In 1893 Carl Barus was also the first to determine the melting curve of basalt as a function of pressure, which enabled King to propose a geothermal gradient for the Earth.

In 1909 the British geologist Alfred Harker summed up his opinion on magma generation in The Natural History of Igneous Rocks: “We must seek the immediate cause of igneous action, not in the generation of heat, but chiefly in relief of pressure in certain deep-seated parts of the crust where solid and molten rock are approximately in thermal equilibrium. We are thus led by an independent line of reasoning to the principle already enunciated, which connects igneous action primarily with crustal stresses, and so secondarily with crust-movements.” Furthermore, “at a sufficient depth, such conditions of temperature prevail that solid and liquid rock are in approximate thermal equilibrium. Any local relief of pressure within that region, connected with a redistribution of stress in the crust, must then give rise to melting.” These statements were an incisive expression of the fundamental principle of melting, but what was still lacking was the mechanism that brought about the relief of pressure.

Not all geologists were ready to embrace this as a solution to the origin of magma, because a viable geologic process that could bring about a pressure decrease was not known. The American geologist Reginald Ald-worth Daly stated (1933): “Local relief of pressure is hopelessly inadequate and need not be further discussed.” Similarly, according to S. James Shand (1949), this pressure effect on melting of silicates was “a quantity scarcely large enough to have any important petrological consequences.”

The progress made in understanding melting and the internal constitution of the Earth by the first part of the 20th century was truly profound, but it had created a paradox. All the evidence was in favor of a solid interior or an exceedingly thick crust at any rate, yet there was a need to account for the magmas erupted from volcanoes. Experimentalists and geophysicists had discovered a process by which melting could occur simply by decompression, but the geologists were unable to find an acceptable decompression mechanism to produce this melting.

Stirring the Pot: Convection

The discovery of radioactivity caused a great stir among scientists, who quickly applied it to an understanding of the Earth. When Pierre Curie discovered in 1903 that radioactive materials emit heat, the Irish physicist John Joly was the first to point out that the abundance and distribution of radium and other radioactive elements may determine the terrestrial heat. In 1910 the British geologist Arthur Holmes (1890-1965) began a study of the natural radioactivity of rocks, and thus began a career that would make a major contribution to both understanding of heat distribution and melting of the Earth by convective flow of solid rocks toward the surface. By 1915, Holmes had calculated a temperature profile for the Earth based on radioactive generation of heat.

After radioactivity was discovered as an internal heat source, it became even more important to establish the principal means of heat loss from the planet. How is this great flux of heat transferred from the deep Earth toward the surface? Holmes proposed that convection was the most effective mechanism to transport heat from depth to the surface. It was an idea that dates back to Osmond Fisher (1881), who proposed, in the days of an Earth with a totally molten interior, that there were convection currents rising under the ocean and descending beneath the continents. In 1928 Holmes proposed that the excess heat is discharged from the Earth by circulation of material in the solid substratum, or the Earth’s mantle, forming thermal convection currents. Convection, and in turn continental drift, was seen as a mechanism to rid the Earth of the great heat generated by the radioactive natural reactor. He considered this substratum to be a solid peridotite rock, which could flow like a fluid at a rate plausible for a deep Earth process—about 5 cm/year. Holmes’s scheme included downward currents below geosynclines, the great sedimentary troughs that today we recognize as subduction zones, and ascending currents below midoceanic “swells,” where “a discharge of a great deal of heat” occurs due to upwelling of peridotite mantle, decompression, and melting. Holmes’s views in 1928 were amazingly modern and laid the foundations for the concept of plate tectonics, although this has not been generally recognized by Earth scientists.

The concept of plate tectonics had a long gestation period, and it began with the realization in the 17th century that volcanoes and earthquakes are distributed in great linear belts over the Earth. This systematic trend of volcanoes and earthquake zones, and the arrangement of the continents, eventually led to the theory known as continental drift in the early 20th century. The earliest map showing the global distribution of volcanoes was drawn by Athanasius Kircher in 1665. With the spread of geographic knowledge, a more refined picture gradually emerged, one of the systematic, linear or curvilinear arrangement of volcanoes on the surface of the globe. Ideas on the relationship between volcanic activity and major structures and tectonics date to Alexander von Humboldt (1822), who pointed out that the linear arrangement of volcanoes on the Earth was proof that the mechanism that generates volcanism was a deep-seated feature. Robert Mallet (1858) compiled the first map showing the global distribution of both earthquakes and volcanic eruptions in a “vast loop or band round the Pacific,” indicating a strong coincidence in their geographic distribution. He also recognized that most of the Earth, especially the interior of the continents, is seemingly free of earthquakes and volcanic activity. He divided the Earth into a series of basins, separated by “girdling ridges,” which are present even on the ocean floor: “It is along these girdling ridges, whether mountain-ranges or mere continuous swelling elevations of the solid, which divide these basins beneath the ocean surface one from the other, that all the volcanoes known to exist upon the earth’s surface are found, dotted along these ridges or crests in an unequal and uncertain manner.” He regarded these ridges as “suboceanic linear volcanic ranges” which mark “the great lines of fracture of the earth’s crust.” Mallet’s recognition of the distribution of these geological structures, which are now basic to the theory of plate tectonics, was probably the earliest.

The recognition of the linearity of volcanoes and earthquake zones coincided with a mobilistic view of the Earth toward the end of the century, a view based both on geophysical reasoning and geologic evidence. Osmond Fisher had argued in 1881 that cooling of the Earth could not be brought about solely by the process of conduction, as had been proposed by Kelvin, but was in part brought about by convection currents in a plastic substratum. In Fisher’s view, these convection currents led to lateral movement of the overlying crust, accounting for much of the character of the Earth’s surface, including linear arrangement of volcanoes, earthquakes, and mountain chains. The ideas of a mobile Earth were first crystallized into a coherent theory of continental drift in 1912 by Alfred Wegener. One of those who sided with Wegener was Arthur Holmes, who greatly strengthened the theory by proposing in 1928 a more plausible thermal convection mechanism for crustal movement than Wegener had put forward. Holmes’s mechanism involved the flow or upwelling of the Earth’s mantle, providing a process whereby decompression of the hot mantle would occur, leading to melting and volcanism. Continental drift was an inevitable consequence of this circulation, but the process was much wider in scope: It involved the lateral motion of the oceanic crust as well. By 1931 Holmes had constructed a complete theory of continental drift, sea-floor spreading and subduction, based on this mechanism.

Holmes differentiated between a largely granitic crust in the continents and a basaltic crust in the oceans. He considered the mantle beneath the oceanic crust as most likely of peridotite composition, and argued that although it was probably crystalline and thus highly rigid, it had some of the properties of a fluid in the context of its large-scale dimensions. He proposed that the convection cells have dimensions at the scale of the mantle, some 2900 km in depth, and that their ascending limbs rise beneath the oceans, where they generate oceanic “swells” analogous to our modern concept of midoceanic ridges. Ascending currents he proposed were accompanied by decompression and partial melting of peridotite, to form basaltic magma, giving rise to volcanism. Descending currents were caused in the mantle by sinking of high-density rocks. The process he outlined was essentially the same as the modern concept of subduction: The convergence of crustal plates on the Earth leads to underthrusting of one plate and its descent deep into the mantle. Holmes estimated that the velocities of the convection current were on the order of 5 cm/ year, which, as was discovered in the 1960s, is perfectly within the range of typical sea-floor spreading rates. Although widely accepted in Great Britain, Holmes’s ideas on mantle convection and a mobile Earth were generally not embraced elsewhere for some time. When the idea of sea-floor spreading was finally accepted by a majority of Earth scientists in the 1960s Holmes’s fundamental contribution was frequently ignored. One is reminded of the chilling words of Sir William Osler: “In science the credit goes to the man who convinces the world, not to the man to whom the idea first came.” Is it more important in science to convince than to discover?

The discovery of plate tectonics has now provided the long-sought mechanism for decompression in the Earth’s mantle—the most important process that brings about melting and generation of magmas. It is clearly the mechanism that generates magmas below the midocean ridges of the Earth and is thus responsible for most of our planet’s volcanism.

Birth of Petrology

Among the earliest observations on the mineral content or petrography of volcanic rocks were Leopold von Buch’s studies in 1799 on volcanic rocks near Rome. Von Buch concluded that the crystals of leucite had formed while the lava was still fluid, and discounted any hypothesis to the effect that they had been precipitated from an aqueous solution and later incorporated in the magma. Basalts and other volcanic rocks, however, are composed of exceedingly small crystals, which are invisible with the naked eye or even through a magnifying glass. The pioneer geologists were therefore unable to investigate the internal structure of these rocks and identify their tiny crystal components. The microscopic study of volcanic rocks was first carried out by Pierre Louis Cordier in 1815, who crushed basalts and other volcanic rocks to a fine powder, separated the various particles by a flotation process, and examined them by microscopic and chemical tests. He concluded that ancient basalts were very much like modern volcanic rocks in texture, mineralogy, and other characteristics—an important step in solving the Neptunist versus Plutonist controversy over the origin of basalts.

The art of making thin sections began with the sectioning of fossil wood, but the method was soon applied to rocks. They could be sliced so thin as to be transparent, and their minerals could be identified and their textures studied in detail. By 1829 the polarizing microscope had been invented by William Nicol and was used widely for the study of crystals. Henry Clifton Sorby (1826-1908) was the first geologist to make thin sections of volcanic and other igneous rocks for microscopic study.

Another crucial step in the study of volcanic rocks was the development of chemical analysis. Studies of the chemical composition of volcanic rocks date back to the work of Abbé Lazzaro Spallanzani (1794), who reported on the chemical composition of volcanic rocks in Italy in terms of weight percent of the five major oxides of silica, alumina, lime, magnesia, and iron. In 1805 the Scottish chemist Robert Kennedy reported chemical analyses of whinstone or basaltic rocks. He was able to dissolve basalt by melting the rock powder with caustic potash at high temperature in a silver crucible. He then showed that it contained 48% silica, 16% alumina, 9% lime, 16% iron oxide. Water and other volatile components accounted for about 5%. About 6%, then, was not accounted for. Kennedy suspected this was in part “a saline substance,” and, after an elaborate analytical routine, showed that it was about 4% soda. He thus accounted for 99% of the chemical constituents of the rock—a great feat at the time.

When the early geologists began to accumulate data on volcanic rocks, they discovered that the chemical composition of the rocks, even those from the same volcano, varied greatly. Confronted with the problem of how to account for this diversity, George Poulett Scrope proposed in 1825 that all types of igneous rocks were formed from a single parent magma, which gave rise to a variety of types through a process of differentiation before crystallization. It was Charles Darwin, however, who laid the foundation for the process of “magmatic differentiation,” through his discovery of crystal settling due to density differences between crystals and the magmatic liquid. During his exploration of the volcanic Galapagos archipelago in 1835, Darwin studied a basaltic lava flow on James Island (now San Salvador). He noted that the crystals of feldspar were much more abundant in the base of the lava than in the upper part, and deduced “that the crystals sink from their weight.” Von Buch had earlier (1818) noted a similar concentration of crystals in the lower part of obsidian lava flows on Tenerife in the Atlantic, interpreting this as the settling of crystals in the lava during and after flow. In his Geological Observations on The Volcanic Islands, Darwin considered this observation as “throwing light on the separation of the high silica versus low silica series of rocks.” He realized that once early formed crystals sink, the remaining magma would be chemically different from the initial liquid.

Another process was developed by Robert Bunsen in 1851 to account for the chemical range of volcanic rocks, after his discovery of two magmas in Iceland, with radically different chemical composition. Bunsen was struck by the abundance of two volcanic rock types: yellow and multicolored, silica-rich volcanic rocks (rhyolite), and the dark-colored basaltic rocks. A few volcanic rocks were of intermediate composition, and he proposed that they were the result of mixing of the silcic and basic magmas. Bunsen’s contribution was to put forward the first viable hypothesis that could account for the diversity of the chemical composition of volcanic rocks by the mixing of two primary magma: basalt and rhyolite. Traveling with Bunsen in Iceland was Sartorius von Waltershausen. He was the first to discuss the distribution and occurrence of the trace elements, as well as the major oxides in volcanic rocks, and listed 23 trace elements.

In the 18th century the volcanic fluid had become known as magma, derived from the Greek word which refers to a plastic mass, such as a paste of a solid or liquid matter. The term magma was used early on in pharmacy as in “magnesia magma,” or milk of magnesia, and the word passed from pharmacy into chemistry to represent a pasty or semifluid mixture. From chemistry it passed into petrology to replace “subterraneous lava.” What was the nature of this fluid inside the Earth? Bunsen, in Uber die Bildung der Granites, stressed that magmas are nothing more than a high-temperature solution of different silicates: “No chemist would think of assuming that a solution ceases to be a solution when it is heated to 200, 300 or 400 degrees, or when it reaches a temperature at which it begins to glow, or to be a molten fluid.” This was a fundamental breakthrough in thinking about crystallization from magmas. Bunsen further recognized that minerals crystallized from magmas at a much lower temperature than the melting point of the pure minerals. If magmas are true solutions, then they should be able to mix, as indeed Bunsen had proposed on the basis of his Iceland work.

During the latter part of the 19th century the chemical analysis of volcanic rocks became commonplace. Leading petrologists considered that most igneous rocks were derived from basaltic magma by a process of differentiation, but no one researcher contributed more to the understanding of the processes that account for the chemical diversity of igneous rocks than the American petrologist Norman L. Bowen. From high-temperature experiments, Bowen discovered that magma could greatly change in composition as it cooled, if the early formed crystals were effectively separated from the liquid, thus providing proof for the fractional crystallization process first proposed by Darwin. Bowen was a strong advocate for fractional crystallization as the primary process responsible for differentiation, and considered that basaltic magma was the primary or parent magma of all volcanic rocks. The new understanding of magmatic differentiation was made possible by Bunsen’s recognition that magmas were true solutions, Darwin’s concept that settling of crystals resulted in a change in the composition of the liquid, and Bowen’s systematic experiments on crystallization of a primary basaltic liquid. But how, then, was this primary basaltic liquid formed?

The Source

Throughout the first half of the 20th century, two principal hypotheses existed on the nature of the source of magmas. Because the principal type of magma erupted was basalt, it appeared logical to some that the source region was also basaltic in composition, and possibly the high-pressure form of basalt known as eclogite. The basaltic magma would then be derived by wholesale melting of the source. The other view was that the magma was derived by partial melting of peridotite. It had long been noted that xenoliths of both peridotite and eclogite were ejected from some volcanoes. Because they were brought up in the magma, it seemed logical that these rock types might be an indication of the source region. As early as 1879 it was suggested that peridotite was the source of basaltic magmas, and by 1916 Arthur Holmes proposed that magma was generated by melting of peridotite at a depth of several hundred kilometers.

At the beginning of the 20th century, geophysical measurements were giving results in favor of a peridotite source. This was done by measuring in the laboratory the physical properties of a number of rock types under high pressure and temperature. By comparing these results with the speed of earthquake waves traveling through the deep Earth, it was shown that properties of the Earth’s interior are comparable to those of peridotite and other ultrabasic rocks under great pressure in the laboratory. Geophysicists therefore generally adopted a peridotite composition for the mantle.

Bowen (1928) also proposed that the only rock type that could give rise to basaltic magma by partial melting is peridotite, and suggested that this occurs at depths of 75-100 km. Bowen then addressed the question of whether the partial melting occurred as a result of re-heating or due to pressure release. Arthur Holmes’s ideas on convective flow in the mantle were published in the same year as Bowen’s classic work, but Bowen seemed unaware of them. Having no plausible mechanism to bring about a pressure release, he stated: “It seems necessary to leave open the question whether selective fusion takes place as a result of release of pressure or as a result of reheating.” He therefore had considerable difficulty in developing a scenario of pressure-release melting in the peridotite mantle. Only with the development of geodynamics and acceptance of Holmes’s theory of mantle convection could it be shown that a peridotite mantle can rise to regions of lower pressure, thus bringing about melting without the addition of heat.

The theory of partial melting of peridotite was still debated and considered on a weak foundation by many in the first half of the 20th century. Another group of geologists adopted a totally different view, and maintained that basaltic magma was derived from a deep melted layer or from the complete melting of a basaltic layer at depth in the Earth. This view persisted with some until the 1960s. In an influential book on volcanoes, Alfred Rittmann proposed in 1936 that all volcanic rocks were ultimately derived from a global layer of basaltic magma. Rittmann’s views were adopted by the petrologist Tom F. W. Barth in 1951: “The fact that wherever the crust is deeply rifted, be it in continents, oceans, or geosynclines, basaltic magma is available and capable of invasion is a proof of the existence of a subcrustal basaltic magma stratum.” He was fully aware of the objections of geophysicists, that such a layer could not transmit some seismic waves, contrary to observations, and proposed that the basaltic magma behaved like glass at the high pressures beneath the Earth’s crust and was thus able to transmit these waves. In his 1962 work on Theoretical Petrology Barth stated: “It is necessary, therefore, that the basaltic substratum is in a molten (noncrystalline) state at its subterranean locale. Otherwise it would not reach the surface with a homogeneous composition as we actually observe.”

The American geologists Francis J. Turner and John Verhoogen accepted the seismic evidence for a solid character of the Earth’s mantle in 1951, and considered it composed either of peridotite or of eclogite. From this it followed that primary basaltic magma had to be derived from a largely solid mantle, whose composition was other than basalt. The observed temperature increase with depth and heat generation by radioactive decay within the Earth could produce the required heat source. In their classic textbook on Igneous and Metamorphic Petrology, Turner and Verhoogen stated: “The problem of the origin of basaltic magma is thus not so much that of finding adequate heat sources as that of explaining how relatively small amounts of heat may become locally concentrated to produce relatively small pockets of liquid in an otherwise crystalline mantle.” In what seemed like a fleeting moment, Turner and Verhoogen considered the hypothesis that basaltic magma could be produced by melting of a peridotite mantle due to pressure release, but then quickly rejected it: “This hypothesis of magma generation is possibly the most popular one, although it is unacceptable to the present writers. It is difficult indeed to see how pressure can effectively be reduced at such depths.” They realized of course, that melting could occur if convective transport of mantle material occurred to shallower levels, as originally proposed by Arthur Holmes. This was clearly a viable mechanism as long as the deep mantle temperature was greater than the melting point of the mantle at shallower depth. Yet they were not ready to accept the existence of convection in the solid Earth, which led them to conclude: “Whether convection does occur in the mantle is not definitely known, nor is it known whether it could be effective in the upper part of the mantle where magmas are generated. Thus, although convection might lead to melting, it cannot be shown that it does, and the problem of the generation of magma remains as baffling as ever.”

This statement is a good reflection of the state of affairs regarding theories of magma generation at the middle of the 20th century. While great strides had been made in understanding the chemistry and mineralogy of volcanic rocks, the Earth scientists were at a loss to explain melting. This was due almost entirely to the static view taken of the Earth’s interior at the time—it seemed inconceivable to most that the rigid and solid mantle could convect. However, the discoveries made on the ocean floor in the 1960s and development of geodynamics were to change all that. Although melting was poorly understood at the time, knowledge of the mantle source region was advancing through studies of xenoliths. A great number of xenoliths are of peridotite composition, giving further credence to the idea that the source region of basalt magma, the Earth’s mantle, is dominantly composed of peridotite.

The Role of Water

Great clouds of gases and steam emitted by volcanoes were long considered to be derived from groundwater, surface waters, such as nearby lakes or streams, or sea-water. This was evident, according to many scholars, from the location of volcanoes near the ocean or on islands. The role of water was thought to be crucial in generating explosive eruptions and to have an important effect on the viscosity of magmas. In 1794 Spallanzani recognized that several gases are important in lavas and volcanic regions, including “hydrogenous gas, sulfurated hydrogenous gas, carbonic acid gas, sulfurous acid gas, azotic gas.” But another, more powerful agent, he noted, was “water, principally that of the sea,” which communicates by passages with the roots of volcanoes. On reaching the subterranean fires it suddenly turns to vapor, and the elastic gas expands rapidly, causing volcanic explosions. Supporting his hypothesis, on a more practical level, Spallanzani cited accidents in glass-making factories, in which molten glass was poured into molds not completely dry or free of water, causing dreadful steam explosions.

Scrope (1825) attributed the fluidity of magmas to water: “There can be little doubt that the main agent … consists in the expansive force of elastic fluids struggling to effect their escape from the interior of a subterranean mass of lava, or earths in a state of liquification at an intense heat.” These “elastic fluids” Scrope considered to be mainly steam and other volcanic gases. It was the expansion of gas in the magma that led to its rise in the Earth’s crust, and led to violent and explosive eruption upon reaching the surface.

Scrope discussed in some detail the evolution of steam in magmas at depth, and was perhaps the first to point out that under great pressure, water would be dissolved in the melt, but upon decrease in pressure or increase in temperature the water would be vaporized, leading to explosive eruption. From the increase of temperature with depth in mines, Scrope concluded that at great depth, the Earth was at an intense heat, and that this great accumulation of caloric in the deep Earth led to continued flow of caloric toward the surface, in order for the heat to attempt to attain equilibrium. Thus Scrope considered the formation of magma as due to the passage of caloric by conduction from depth to upper levels in the Earth, where the caloric led to melting of rocks. Scrope proposed that the caloric in molten rocks was in large measure due to water: “There is every reason to believe this fluid to be no other than the vapour of water, intimately combined with the mineral constituents of the lava, and volatilized by the intense temperature to which it is exposed when circumstances occur which permit its expansion.” The water of the oceans, he theorized, was derived from the interior of the Earth due to degassing through volcanoes. Large quantities of water “remain entangled interstitially in the condensed matter” in the deep-seated rocks, and the escape of steam during volcanic eruptions carries off an immense amount of caloric, leading to rapid cooling and consolidation of magmas at the surface.

By 1865 the French geologist M. Fourque had measured the amount of water in the volcanic products of Etna and estimated the amount of steam discharged from the vent. In the latter part of the 19th century geologists began to develop more sophisticated models, proposing that the steam emitted by volcanic eruptions was primordial or of deep-seated origin, rather than just recycled surface waters. Among them was Osmond Fisher (1881), who regarded volcanic gases as original constituents of magma. The potential role of escape of water from volcanoes in the formation of the oceans began to surface as an important hypothesis. The German geologist Eduard Suess proposed that all water in the oceans and atmosphere came from the outgassing of the interior of the Earth. The concept of volcanic recycling of water from the ocean to the atmosphere and back again to the deep Earth was proposed by John Judd (1881), who pointed out that volcanoes are generally located near the ocean, and speculated that fissures may transport seawater from the ocean to the magma at depth. “Volcanic outbursts” could be due to water finding its way down to a highly heated rock mass, lowering the melting temperature and causing melting. Judd considered high temperature and the presence of water and gas as the key factors of volcanic action and argued that magmas can absorb or dissolve large quantities of water, which escapes violently during explosive volcanic eruption, as Spallanzani had first pointed out. The magmas could absorb water or gases either initially, as primordial gases during formation of the globe, or at any stage in geologic history, due to infiltration of water into the Earth’s crust.

With the widespread recognition of water in magmas, it was logical to consider its role in explosive eruptions. In the age of steam during the Industrial Revolution of the 19th century, John Judd pointed out that “a volcano is a kind of great natural steam-engine.” Bonney (1899) also compared volcanoes to a boiler, emphasizing that the steam in magma is the main explosive force in an eruption. He noted that the volume of steam is nearly 1700 times that occupied in the form of water, an enormous expansive force, adequate to account for volcanic explosions. The origin of volcanic water, according to Bonney, is related to the proximity of volcanoes to the ocean and the percolation of rainwater into the magma. He stressed the importance of addition of water to lower the melting point of rocks, and followed Osmond Fisher and others in attributing eruption to the presence of water. When water is depleted or withdrawn from the system, the eruption ceases.

In the latter part of the 18th century, Dolomieu had determined that many deposits in Italy and on Sicily, which at first sight looked like normal stratified sediments, were in fact consolidated volcanic ash deposits, the products of fallout from explosive eruptions. The disruption of magma as it is blown into the air leads to fragmentation and the formation of volcanic ash, tephra, scoria, and other forms of pyroclastic material. We now know that the primary agent of this disruption is the explosive expansion of steam. Sartorius von Waltershausen (1853), a true pioneer in the study of pyroclastic rocks, was the first to attribute their formation to the effects of water on the magma. He proposed that magma rises and erupts due to pressure of water vapor escaping from the magma and that the volume increase of water vapor was also responsible for the fragmenting process—the formation and ejection of pyroclasts such as pumice and ash. He also recognized that peculiar volcanic rock deposits can result when magma enters the ocean or is erupted under water. It was during travels in Sicily in 1835 that he first came across a brown tuff, a homogenous rock, composed largely of a single mineral. He named the rock palagonite after the nearby town of Palagonia, and by chemical analysis determined it as one unusually rich in water and iron, with about 12-23% water. He became curious as to the rock’s origin, and noting that it was often associated with marine deposits, proposed that it was a volcanic product that forms thick layers in many submarine volcanic formations. In 1846 he and Robert Bunsen studied palagonite mountains in Iceland, and noted that a zone of palagonite tuffs stretches across Iceland. Often in association with palagonite he noted, was a black, water-free and glass-like material, similar to obsidian, which he gave the name sideromelan. He was able to show that palagonite is basically sideromelane or basaltic glass that has taken up water, and from its geologic setting, that palagonite is a product of shallow subaqueous or submarine basaltic eruptions. Von Waltershausen was correct in this deduction; we now know that the Icelandic palagonite rocks are formed by volcanic eruptions below the thick ice cap that covered Iceland during the last ice age, and that the Italian palagonites were erupted in the ocean.

Bunsen, on the other hand, disagreed with von Waltershausen, considering the palagonite tuffs as the products of basaltic rocks metamorphosed in the presence of much water and carbonate. He showed by chemical analysis that the tuffs were virtually identical to basalt lava after the high water content had been subtracted. His study was not restricted to Icelandic rocks, because Charles Darwin had given him samples of palagonite from the Cape Verde islands.

Another volcanic rock of basaltic composition is lava-like but composed of rounded or pillow-like forms. The origin of this rock does not have a bearing on water in magmas, but rather on magma in the water. The identification of pillow lava dates back at least to the 1870s. Later it was proposed on the basis of observations in Italy that it forms due to submarine eruption. The British geologist Tempest Anderson observed an eruption in Samoa, noting that when the basaltic lava was flowing into the sea, the submarine component of the lava formed bulbous masses and lobes with the shape of pillows. By 1914 a subaqueous origin for pillow lavas had been established. They were already known to be present in the geologic record in association with marine sediments in Scotland, and also as products of subglacial eruptions, such as in Iceland. Oceanographers discovered only in the 1960s that basaltic pillow lava forms the floor of most of the world’s oceans and is thus the Earth’s most abundant volcanic rock but also the most remote for study.

When high-pressure melting experiments were first carried out at the beginning of the 20th century, it became evident that magmas residing deep in the Earth can contain much more water in solution, and that this water must be liberated when they are erupted at the surface. In 1903 C. Doelter was one of the first to propose that magmas at great pressure in the Earth’s crust may dissolve water and that the magmas may become explosive when they reach the surface of the Earth. In support of this theory, the French geologist Armand Gautier carried out laboratory experiments on volcanic activity in 1906, suggesting that magma rises in the crust as a result of gas expansion, and attributed the violence of volcanic eruptions to the explosive liberation of water from the magmas. By the early 20th century the fundamental ideas about the causes of explosive volcanism and the importance of water had thus been firmly established.