By Köppen’s account, the new director after 1903 had no understanding of scientific work and took his task at the observatory to be the bringing of order and system to its affairs. This meant much filling out of cumbersome forms, clock punching, and deadline meeting, none of it, as Köppen said, very compatible with scientific work. Moreover, the admiral seemed not to care what sort of work was done as long as the forms were filled out and the deadlines met. Köppen managed to get his scientific work done anyway because he had finally obtained funds for his meteorological kite and balloon station at Großborstel. It was at this time that Köppen moved his family out of Hamburg to the big house on the Violastraße. He was able to convince the admiral to allow him to work at the station two days a week and come in to Hamburg and the observatory on four days.7
However much Alfred hated administrative work—and he did—he would at least have two weekdays, his Sunday, and his evenings for his research, as had Köppen. He did not deliberate long. Already, by the first week of January 1919, he had decided. He wrote to Köppen that he had been to Berlin, to meet with various officials and discuss his future. “With Marburg,” he wrote, “it is now a definite no. At present it is impossible to create any new positions.… B. and K. advised me, off the record, to take the position at the Observatory.”8 “B and K” were, respectively, Carl Becker and Kohlschütter. Becker was the official at the Prussian Ministry of Culture who had approved the short-lived professorship for Wegener at Dorpat. It appears that Kohlschütter, during the Berlin meeting, virtually assured Wegener that he would appoint brother Kurt to be the head of the Weather Forecast Division (Division III) at the observatory.
So, wrote Else, “for us the die was cast.”9 Wegener accepted the appointment as meteorologist of the observatory, effective 1 April 1919 (though not due until September 1919). He was touched by the regret of his Marburg colleagues: “I had the feeling, for the first time this time, that they really wanted to keep me at the University and just were unable to do so. The Faculty seems really to have put their shoulders to the wheel.”10 He would remain at the university in Marburg until the end of the winter term, and then he, Else, and the children would move to Hamburg in the summer. Else was delighted; she had wanted this as early as 1912 and in some sense had never wanted to leave home at all, but only to bring her husband to her family home. This she had now almost accomplished.
The Köppens offered to give up their house entirely and find a new, smaller lodging in the neighborhood, but Alfred and Else refused; it was just too much to ask. They would expand the house, and in the meantime they would take the ground floor, while the Köppens senior moved upstairs. This would be somewhat inconvenient, if only because Alfred required a dedicated workroom away from children, but they could not count on having enough money to expand the house in any case; Kurt predicted that the Central Bank in Berlin would fail as early as March. Berlin was in the midst of the Communist revolution known as the Spartakus uprising, and there was street fighting between the Communist revolutionaries, led by Rosa Luxemburg, and the anti-Communist Freicorps. Alfred began to think he should get his family to Hamburg soon, in case the predicted collapse of the bank turned their currency and savings into waste paper.
Alfred’s previous reluctance to move to Hamburg had vanished. He was not, as he would have been in 1912, a temporary, part-time assistant writing for hire; he was now to be the successor (Nachfolger) to one of the greatest meteorologists of the era, as well as a salaried and pensioned senior scientist at a major and respected research institution. He would live in a house that was a pilgrimage site for every meteorologist who came to Germany, in which he had been an awed guest in 1908. He sketched out enthusiastic plans. Kurt would come and head the Weather Division; Alfred was certain that Kurt was capable of making the kite observations himself even though he had not pursued forecast meteorology since early in the war.11 Alfred envisioned expanding the kite station to include a runway for fixed-wing aircraft, as Kurt was convinced that this mode of observation was the wave of the future. “How strange it would be,” he mused, “if the brothers Wegener were to come together once more now, as they were in Lindenberg.”12
Meteors and the Moon
Before Hamburg could be a reality, there was still the winter semester in Marburg, to which he was already committed. He was also extremely anxious to restart his scientific career, so many times interrupted in the war years. In December 1918 and January 1919 he was able to discover that the last of his publications from the war years had finally appeared. These included a digest in the Astronomische Nachrichten of his successful plan to find the Treysa meteorite, a similar short summary of his book on tornadoes and waterspouts in the Meteorologische Zeitschrift, and a highly mathematical treatment of mirages, on which he had worked intermittently over the past three years, in the Annalen der Physik.13 As he had promised himself and Köppen, he was determined not to bury his work any longer in obscure journals. This was very much the case for his greatly expanded treatment of color change in meteors, which he published in the Acta of the Royal German Academy of Sciences.14
When he finally saw the published work on color changes of large meteors, he was very pleased with the way his ample (thirty-four-page) treatment had turned out. It was the initial article in volume 104 of the Acta and thus very likely to be read. He had followed the conservative and cautious template he had established with his work on tornadoes and waterspouts, providing a detailed descriptive catalogue of all reliable accounts of the color changes in large meteorites, going back through the nineteenth century.
His catalog recorded the measured elevation, angle of descent, color of the fireball, and color (if any) of the Rauchschweif (smoke trail), especially in those rare instances confirmed by spectrographic analysis. He had then aggregated these data, to show that the color change was almost invariably from green to red, and that this shift could be explained by reference to the composition of the atmospheric layer through which the meteor was passing. He then illustrated this with his own diagram (first published in his Thermodynamik der Atmosphäre) of the elemental segregation of the atmosphere in the sequence hydrogen→nitrogen→oxygen and correlated the shift in color with the passage from a predominantly hydrogen to a predominantly nitrogen atmosphere. He put forward the opinion, largely confirmed by later observers, that the color of the incandescent trail of a meteor passing through the atmosphere is mostly the result of the ionization of the air and only slightly the result of the meteor’s own composition.
As often happened in reviewing his own published work, new ideas occurred to him, and aspects of the topic emerged that were not at first apparent. He saw that there was something more to say about air resistance. He had calculated the probable diameters of meteors producing green and red spectral signatures. He knew that iron meteorites tended to burn up more rapidly in the atmosphere, but he was curious why more large stony meteorites were not discovered, given the large size (more than 100 meters [328 feet] in diameter) estimated for a number of his catalogued objects. One notes that objects in this size range are today denominated as asteroids, with meteoroids restricted to smaller diameters; this discrimination was not yet established in the early twentieth century.
In thinking about air resistance, which caused the frictional heating of the face of the descending body, he had to consider the extent to which the material on the leading edge would not only be heated but also be compressed until the resistance of the air overcame the strength of the stony material and caused it to disintegrate. This would explain the repeated fireball puffs in a descending smoke trail, as the leading face of the meteoroid underwent a resistance greater than its strength and disintegrated. This repeated process of heating and disintegration in the atmosphere would explain why so few meteorites were discovered, and it also explained the survival of his Treysa iron meteorite from 1917, which had landed intact as a 63-kilogram (139-pound) mass. The cohesive strength of iron is much greater than that of rock, and an iron meteorite
impacting a solid rock surface could pulverize that surface without itself being destroyed. Such considerations of solid mechanics had been important for him in his work on continental displacements, and his phrase “molar forces overcome molecular forces” here had a parallel, in which the resistance of the air overcame the molecular cohesion of the heated body.15
This study of meteors, their trails, and their occasional impacts on Earth’s surface led him to consider another related problem: meteor impacts on the Moon. A planetary body with no atmosphere would be subject to bombardment by large meteors that would impact at very high velocities (up to 80 kilometers per second [~18,000 miles per hour]) and likely excavate large craters. There were various hypotheses about the origin of lunar craters: that they were the remnants of exploding bubbles of gas rising in an original magma, that they were the result of “tidal overflow” of lunar magma through fissures in the surface, that they were volcanoes, and finally that they were the result of meteoric impacts. While both the bubble hypothesis and the tidal hypothesis had had strong advocates, the leading candidate was the volcanic hypothesis, with the infall (Aufsturz) or impact hypothesis a distant fourth.16
The Moon was constantly reappearing in his work. Each of the major geophysical theories about motion of continents (proceeding his own) had also contained a hypothesis of the origin of the Moon.17 He had just finished (in the fall of 1918) reviewing the lunar observations of Koch in Greenland during 1906–1908 (finally published in 1917), measurements that he had hoped would provide evidence for the westward movement of Greenland. Additionally, he had completed and published, within the previous two years, an important and widely read paper on lunar tides in the atmosphere. Lunar tides had also figured prominently as one of the several speculative mechanisms for lateral displacement of the continents in both his papers of 1912 and his 1915 book on the origin of continents and oceans.
As if these considerations were not enough, interest in the Moon had burgeoned in the first two decades of the twentieth century. “Selenology” was the name given to the study of the origin of the Moon, the mineralogy of its surface, and the interpretation of its surface morphology. In the nineteenth century these considerations were the province of “selenographers”—those telescopic observers who devoted their careers to exactly describing and recording the features of the Moon’s surface. By the early twentieth century, however, the most prominent figures among selenologists were geologists. Leading the way to a fuller consideration of the surface features of the Moon, geologists developed the analogy between the lunar craters and those forms on Earth which were most like them in shape: volcanoes. The development of volcanology in the early twentieth century reinforced the long-standing interest in the analogy between lunar craters and terrestrial volcanoes; numerous investigations of volcanoes and volcanic phenomena continually added new empirical content to an old impression of similarity. By 1914, volcano observers had established the range of sizes of volcanic craters on Earth—altitudes, diameters, and depths—and it was also possible to estimate the number of active and extinct volcanoes on Earth’s surface.
In 1909, in the final volume of The Face of the Earth, Eduard Sueß had devoted a chapter to the Moon, in which he endorsed the ideas that the Moon had separated from Earth and that the lunar craters were volcanoes. It is not incidental that Sueß, in the very same chapter, vigorously rejected the idea of isostasy and the idea of a floating crust as incompatible with both his notions of the strength of rocks and his fundamental commitment to the contraction hypothesis. To the extent that Wegener was rewriting Sueß on the face of Earth, he might also rewrite him concerning the face of the Moon; this he had not yet done.
As we have already seen, Wegener took great intellectual pleasure in the process of sorting out areas of study he considered unnecessarily muddled. He repeatedly characterized his own creative work as the reorganization of existing elements into more systematic and meaningful arrays—reorganizations that he had the pleasure to carry out, but that might well have been carried out by others. Typically these began and ended within the confines of the existing literature and the realm of calculation.
His consideration of the infall or impact hypothesis of lunar craters began as such a reordering of existing literature, but it did not end there. Writing about the work he did in the winter of 1918/1919, he remarked that
[in] the literature we can see the gross uncertainty afflicting [previous] conceptualization of the details of the impact process, an uncertainty that is the cause of the many missteps and conflicting opinions on the subject. These follow from our nearly complete lack of experience with the sequence of events an impact sets in motion, and the forms resulting from it. The few experiments that have been conducted to date in this area are completely unsystematic, and hardly serve to support the hypothesis. It was to correct this lamentable state of affairs that I performed in the winter of 1918/1919 at the Physical Institute in Marburg, a systematic series of experiments with impact craters.18
Wegener wanted to pursue a systematic experimental approach to lunar cratering. The principal failure, in his mind, of all the previous attempts to simulate lunar craters in the laboratory was twofold. The first failure was that none of the previous experimenters, back into the beginning of the nineteenth century, had any clear sense of what their experimental simulations were meant to demonstrate beyond the simple indication that something that looked like a lunar crater could be created by throwing or dropping some projectile into a yielding medium. The second failure was the lack of any attempt to quantify the results in physical terms to see whether the results obtained could “scale up” from centimeter and millimeter dimensions to tens or hundreds of kilometers, the latter being the dimensions of the largest known craters on the Moon.
The best example of this failure, the more notable since it was a favored hypothesis, was the idea that lunar craters had been formed by gas explosions in hot, viscous magma, with the ring mountains (the crater rims) as the remnants of exploded bubbles. Part of the appeal of the bubble hypothesis was the ease with which a laboratory simulation could be made with a variety of substances (wax, clay, and pastes of gypsum, lime, or sulfur) in which a viscous mass was heated to boiling and then cooled, or heated and aerated as it cooled; most supporters of the idea had undertaken just such simulations.
In parallel with his experiments on impacts in 1918/1919, Wegener repeated experiments in the recorded literature on the bubble hypothesis and found that the resemblance between the small ringwalls produced in this way and the rims of actual lunar craters was quite superficial. Even more importantly for Wegener, who would return to this point again and again in evaluating the candidate hypotheses, the numerical proportions of craters produced in this way did not approximate those compiled (by careful selenographic mapping) for the craters of the Moon.19
Though it was easy to demonstrate that the bubble crater experiments did not produce a crater morphology like that of the Moon, it was easier still, from Wegener’s standpoint, to show that “the hypothesis is based on a fallacy.”20 In attacking the bubble hypothesis on physical grounds, Wegener understood—in a way he had not in 1915—that while the concepts involved in scaling physical forces in models were easily grasped, they were not part of the working vocabulary of most geologists or astronomers of the time. The hostile, baffled reception of his book on continental drift and the failure of his geological critics to understand isostasy, or to even understand the nature of geophysical arguments, had alerted him to the necessity of spelling out the reasoning behind what was, for him, an obvious physical point.
In the laboratory setting, and on the centimeter scale, one can produce a variety of effects that depend on the action of what Wegener generically called “molecular forces”: electricity, magnetism, cohesive strength, and so on. In working with these forces, it is possible in most cases to ignore the “mass forces” (other than weight); in the laboratory, the gravitational attraction between two objects can scarcely be measure
d even with sensitive instruments. The point Wegener pressed home was that on a cosmic scale, the situation is reversed. When one is dealing with planetary-sized bodies, molecular forces are completely overshadowed. The larger the masses involved, the more powerful are the mass forces, while the molecular forces do not increase at all. The result is that on a cosmic scale, laboratory experiments based on molecular forces do not carry over.21
In the case of the bubble hypothesis, the physical phenomenon involved was that of surface tension (Oberfläschenspannung), and though “it is possible to form bubbles of different sizes … always within definite limits … there cannot be, nor could there ever have been larger bubbles on the Moon.… Those who would explain immense lunar features hundreds of kilometers in diameter as burst bubbles commit a fallacy as outrageous as someone attempting to ascribe the flotation of ocean liners to the surface tension of water, by analogy with water beetles or water striders, or the fact that a needle can float.”22
While the tidal and the bubble hypotheses of lunar craters could be dismissed easily on physical grounds, the volcano hypothesis, which was physically plausible but could not be convincingly modeled in the laboratory, required a much more extensive critique. A range of geologists who were experts on terrestrial volcanoes had long supported the idea that the craters of the Moon were volcanoes. Wegener’s argument against the volcano hypothesis may be summarized thus: though lunar craters may seem to look more like terrestrial volcanoes than like other known geological phenomena, they don’t look much like them. When the two sets of forms are systematically compared, they can be seen to differ in shape, to sit differently with respect to their surroundings, to manifest a very different range of sizes, to be arranged differently relative to one another, and to be distributed quite dissimilarly on the planetary surface.23
Alfred Wegener Page 72
