The Moon’s surface is covered by so many craters that it is doubtful whether there is a single point on it that has not been at some time in the interior of a crater. The volcanoes on the earth, on the other hand, cover an infinitesimal portion of the surface, and their craters are so small that were there astronomers on the Moon, even using the most powerful telescopes they would be able to detect them with great difficulty or not at all—who knows if such lunar astronomers would even be aware of their existence? On the Moon, on the other hand, the crater form is represented at every size up to the gigantic dimensions of the Mare, and are not only the most typical and common surface forms, but, as one can see at a glance, practically the only forms.
I do not understand how anyone comparing the Moon to the terrestrial globe can come to any other conclusion than this: the forms are fundamentally different, therefore their origins must be different as well. The contrast between the Earth and the Moon is so glaring that the next generation will surely laugh at our frantic attempts to establish an identity between them.24
Turning to the impact hypothesis, Wegener noted that such experiments had been performed with a wide variety of materials. Experimenters had thrown canister shot into a viscous paste of lime, gypsum, and cement. Others had spilled knife tips of dextrin (a fine powder used as sizing and adhesive) onto a surface of the same powder. Still others had used lycopodium powder (the extremely fine spores of ground pine, used in pyrotechnics and to pack burn injuries) as the impacting surface and balls of rubber or yarn as the impactors. All of these had produced satisfying imitations of lunar craters, often including the characteristic central peaks these forms exhibit, but all had some manifest implausibility, physical or geological, mixed in. The canister shot experiment (and that with a rubber ball in lycopodium powder) had used very hard but low-velocity impactors, together with a very yielding surface; the materials were too dissimilar. The experimenter with dextrin had the right scaling, but its proponent thought he had demonstrated that the Moon was covered with dust or sand and had been bombarded with dust or sand.25
Wegener’s experiments, on the contrary, are notable for their strict protocol, careful scaling, accurate measurement, and systematic variation. Above all, they are interesting because they represent the only blind experimental trial in the whole literature of impact craters to compare laboratory craters and real lunar craters, numerically and exactly. Wegener decided, in the course of his experimental design, that he would do a blind comparison with measurements of real lunar craters compiled by Hermann Ebert (1861–1913) in 1890 from a multitude of crater measurements by previous observers. Ebert’s measurement data were well known and well respected; Grove Karl Gilbert (1843–1918) had discussed them in a review article on the history of impact cratering experiments in 1892, an article Wegener had much admired and used to guide his own experimental plans.26
Ebert had looked at ninety-two craters with diameters from 13 to 161 kilometers (8–100 miles). He had measured the diameter of the craters, the altitudes of the ringwalls above the impact surfaces, the depths of the crater floors relative to the impact surfaces, the altitudes of the central peaks relative to the impact surfaces, the ratios of the crater depths to the ringwall heights, the ratios of the diameters of the craters to the crater depths, the ratios of the crater depths to the heights of the central peaks, and so on.27 Wegener made note of these various kinds of measurements and ratios but did not examine the actual numerical results of Ebert’s measurements before performing his own experiments.
Wegener used cement powder for both the target surface and the impacting mass. As he explained, “The reason for using such a weakly coherent powder is that since the impacts were produced by hand power alone, the kinetic energy of the impacting bodies was also extraordinarily reduced compared to those striking the Moon. If one wishes to achieve similar results in spite of the reduced kinetic energy, one must proportionately reduce the molecular cohesive strength of the rocks of the target surface, since this is overcome within the limit defined by the borders of the lunar craters: thus we arrive at this kind of powder, with a very small degree of cohesion.”28
This did not, of course, mean that the surface of the Moon was covered with powder, but rather that the powder in the experiment corresponds to the rocky surface of the Moon. Cement was also advantageous because he was able to fix the produced craters by spraying their surfaces carefully with water and then the next day, when they had hardened, soaking the entire mass in water.
After several attempts, he settled on a smooth surface of cement powder in a shallow paper box. He pressed down a centimeter or so of cement powder with a sheet of paper and, removing the paper, sprinkled the surface with about a half centimeter of loose cement. He had discovered that when the surface was too tightly packed, no craters could be produced, and that he had to vary the depth of the impact surface in order to produce central peaks. He produced the craters by throwing a half tablespoon of cement powder, which typically produced craters of 4 or 5 centimeters (1.6–2.0 inches). When satisfied by the general morphology of his craters, he produced eighteen with central peaks. He then measured the diameter of each crater, the altitudes of their ringwalls, the crater floor, and the central peak relative to the undisturbed impact surface. From these values he generated measures of the total crater depth and the total height of the central peak, and using these, he generated additional ratios. He presented these in tabular form, dividing the data into two groups of nine craters each.29
These eighteen craters provided the basis for the comparison with the measurements of real lunar craters by Ebert. While the ratio of diameter to crater depth was off by a good bit from the data of Ebert, Wegener’s ratios of crater depth to ringwall height and of crater depth to central peak height corresponded almost exactly. Wegener subsequently argued that his largest crater, which was 12.2 centimeters (4.8 inches) in diameter, corresponded to a crater approximately 121 kilometers (75 miles) in diameter on the Moon. Ebert had a set of craters up to 161 kilometers (100 miles) in diameter; thus, the loading of Ebert’s data with large craters must make the ratios differ. Wegener was therefore able to argue that he achieved his main objective: the demonstration that by the use of appropriately scaled materials one could produce experimental lunar craters that quite precisely match the dimensional ratios for real lunar craters, and moreover one could do this without introducing selective bias into the experimental trials.30
Outside of this test with preexisting data for the actual Moon, Wegener produced systematic variations in his protocol, changing the material of both the impactor and impact surface (viscous cement pulp) to produce impacts exhibiting double ringwall structures, oblique impacts, crater rows, and furrows, and finally he created a system of rays, some longer than 2 meters (7 feet) in length, by dumping a half tablespoon of gypsum powder on a piece of black pasteboard—thus imitating similar features in the lunar maria.31
In one series of experiments, Wegener used cement powder for the impacted surface and gypsum powder for the impactor, and then when the concrete had set, he cut a cross section through his experimental crater and demonstrated that the impact debris from the impactor spreads all the way to the crater wall, a question of some interest in analyzing the form of the craters.
Wegener’s crater experiments. At the top is a crater produced with an impactor of gypsum onto a surface of cement powder, to show that the debris of the disintegrated impactor would be blown out to help construct the ringwall of the crater. In the middle is a picture of “bipolar ejection” of impact debris, creating a transverse ridge like those observed on the Moon. The bottom photo shows the impact dust streams created by tipping cement onto black paper, creating the structures observed in the lunar maria. From Alfred Wegener, Die Entstehung der Mondkrater (Braunschweig: Friedrich Vieweg & Sohn, 1921).
Wegener emphasized that he had “arrived, following a purely morphological-empirical path, at the result that typical lunar craters are best explained as impact crat
ers.”32 Both qualifications are important. Wegener’s experiments explain only the typical crater form and not the great crater seas and maria, and he was “arguing to the best hypothesis” based on available evidence, restricting the latter to the “purely” empirical (i.e., quantitative) study of crater shapes.
Here as always, Wegener was a strongly visual thinker, and his papers on meteors and the Moon are amply illustrated with drawings and photographs. Objects were, for him, physical processes frozen in time, mathematical law made manifest in three-dimensional, measurable reality—out of the noumenon, into the phenomenon. Wegener took the theory of lunar craters in a direction familiar in his other work, especially that on the origin of continents and oceans: the determination of the origin of a series of forms in terms of a “single genetic principle.”
Wegener gave a talk on his work on air resistance and impacts before the Marburg Natural History Society in February 1919 and published a brief note in their proceedings, which appeared at the end of March.33 He then began work on a longer manuscript for submission to the Royal Academy of Sciences, making his work on impact craters a companion piece to his paper on color change in large meteors which had already appeared in this distinguished journal. He had decided to dedicate this manuscript to Richarz, as a gesture of thanks for all his support over many years.
Hamburg
Even though the winter term at Marburg was not yet over and Wegener was not due to take up his position in Hamburg until September, the political and social instability of Germany in the spring of 1919 forced a change in his plans, and he decided to move as soon as possible. Travel by rail could be accomplished only with police permits for each traveler, and the political situation in March 1919 seemed to be deteriorating rather than getting better. Since the road to Hamburg led through Berlin, time was of the essence. They packed hurriedly and in late March set out for Hamburg, having to carry their own food and an alcohol stove with which to cook it on board the train. It took two days to reach Hamburg, with frequent stops and police searches of the train. Wegener saw them as far as Hannover, the last leg of the trip, and then caught a train back to Marburg; he would live for the rest of the semester in a room at the Physical Institute.34
Even before the move, Alfred had decided not just to get his family to Großborstel but also to take up his position at the Deutsche Seewarte immediately—five months early. While he could not take up his job as Köppen’s replacement until September, he could step in as head of the Weather Forecast Division, replacing Großman, and he indeed took this title (between April and September 1919) and the welcome salary. While Germany was not yet in the grip of the hyperinflation that would beset it between 1921 and 1923, the value of the German mark in 1919 was only half of what it had been before the war; Alfred deemed it unwise and unnecessary for them to expend their slim savings in Marburg while waiting for an appointment to begin in September.
The Köppen-Wegener household in Großborstel, roughly 10 kilometers (6 miles) north of the observatory, was luxurious in comparison with any place that Alfred and Else had ever lived. Surrounded by trees and very close to open country, it had a large fenced yard with berry bushes and fruit trees, as well as an ample kitchen garden. The streetcar line stopped not far from the Violastraße, and from there it was only a few kilometers to the rail station on the main line into Hamburg, and a second streetcar ride to the Seewarte.35 Kurt Wegener made the move to Hamburg at the same time, and Else found him lodgings near the Köppen-Wegener household. Still unmarried, Kurt preferred living in “rooms” rather than a full house of his own.
Beginning in April 1919, Alfred (and Kurt as well) commuted to the Naval Observatory four days a week. The observatory had a number of divisions, but fundamentally the scientific output was split between oceanography/hydrography on one side and meteorology on the other, with this joint preoccupation expressed in the title of the institution’s fifty-year-old publication: Annals of Hydrography and Maritime Meteorology: Journal of Navigation and Oceanography.
The oceanography and hydrography section proceeded with its work much the same way as before the war, though now under the direction of [Carl] Wilhelm Brennecke (1875–1924), who took over general editorship of the Seewarte’s monthly journal in 1919. Brennecke had been at the observatory since 1904 and had been a participant in the German Antarctic Expedition of 1911–1912. Brennecke, Kurt, and Alfred were very close in age and represented the new generation in both meteorology and oceanography. This is exactly what Kohlschütter had wanted: to revitalize the entire institution under the direction of younger men.
Kurt and Alfred were again a team, working together doing practical meteorology. Kurt had been at the University of Straßburg just before the war, in a setting that was pioneering new forecast methods. Alfred had been, as we have seen, a full-time forecast meteorologist since 1916 and had played a role in the modernization of the Domestic Weather Service that year. In the spring of 1919 Kurt and Alfred were able to assemble a staff of men contemporary with themselves, or slightly younger, all able and eager. Kurt recruited Paul Pummerer (1884–1957), trained in both astronomy and meteorology, and who had lost his academic position when the University of Straßburg reverted to French control in November 1918; he was part of that flood of talented men displaced by the settlements of the peace. Alfred brought in his wartime subordinate and favorite coworker Erich Kuhlbrodt (1891–1972), and they were joined by Wegener’s former student from Marburg, Johannes Georgi (1888–1972).
The brothers Wegener decided to begin their tenure at the observatory by completely revising and modernizing all the protocols for collecting, compiling, and disseminating weather data. Forecast meteorology in Germany had suffered in the war years from an inability to get weather data from England and France, those countries immediately to the west and in the path of prevailing weather that would arrive in Germany a day or two later. The observatory had tried to make up for the lack of foreign data by increasing the number of upper-altitude observations; the leading edges of frontal systems (as we would now describe them) are visible first at higher altitudes. This, however, did little to increase the completeness and accuracy of wartime forecasting. Moreover, telegraphic communication between the observatory and its satellite stations was often interrupted—by physical interruption and by censorship during the war, and by strikes and civil unrest during the peace; weather telegrams often arrived in Hamburg too late to have any role in the forecasts.36
One of the first technical changes they introduced at the observatory was a shift away from wire-telegraphic communication to Funkentelegraphie (radio telegraphy), in which messages were transmitted in Morse code on radio frequencies. Not only did this speed up communication, but even while Germany remained out of scientific contact—let alone cooperation—with England, France, and Belgium and had not yet established or reestablished ties with Holland and Denmark, the rapid shift in all of these nations to the transmission of weather data by radio telegraphy in 1919 allowed the Germans to intercept and process the “three-day forecasts” that were the standard in these countries and to employ these data in the preparation of weather forecasts at the observatory. They also engineered a transition in the way forecast data were to be assembled and altered the format and content of the observatory’s weather maps.37
Alfred and Kurt understood that this was an ideal opportunity to reestablish the German Marine Observatory, and especially the aerological station at Großborstel, as a major center of meteorological research and innovation, returning it to the prominence it had enjoyed when Köppen had founded it at the beginning of the century. Köppen’s aerological kite station, constructed in 1903, had burned to the ground in 1913. Though it had been rebuilt shortly thereafter, the recent erection of electrical power lines in the vicinity had made regular kite flying almost impossible; the danger that the kite cables would cut the power lines was simply too great to risk. The station, therefore, was reduced to a daily pilot balloon ascent. Alfred and Kurt would have to
rethink the entire plan of observations at Hamburg.38
In May and June of 1919 Alfred and Kurt drew up extensive plans for a state-of-the-art research institute, to be located a few kilometers to the north of the current kite station and adjacent to the new Hamburg airport, in 1919 already serving civilian domestic aviation. A new kite station, expanded machine shops, a more modern winch, an instrument building, and a repair facility would make this a scaled-down version of Aßmann’s plans for Lindenberg twenty years prior. This new aerological station would, as laid out in detail in Kurt’s 1918 book Vom Fliegen, progressively shift the standard instrument platform away from kites and manned balloons toward fixed-wing aircraft, which could be taken aloft several times a day to record meteorological elements in short flights that could attain readings at altitudes of several kilometers, without the immense expenditure of effort required to obtain the same data with aerological kites.39
So, as Bjerknes had his institute in Bergen and Exner had his institute in Vienna, Alfred Wegener would have his institute in Hamburg. He would push ahead with his plans to advance meteorology as a science, reestablish a new center of gravity for that science in Germany (after the loss of Straßburg), and begin to train a new generation of German meteorologists in modern methods of aerological research. He hoped and expected that the new style of weather maps and the new observation protocols would quickly generate the kind of data needed to understand (in much greater detail) the structure of the troposphere. Rather than accepting the bifurcation of meteorology into a theoretical atmospheric physics on the one hand and practical weather forecasting on the other, he would unify these into a theoretical and practical meteorology that could take its place as an equal among the other sciences in the German university system.
Alfred Wegener Page 73
