1. Fifteen minutes between lectures

Saunders Mac Lane arrived in Göttingen in 1931. In a much later recollection, the American student described an apparently unimportant part of the day. A two-hour class paused for fifteen minutes. Students emerged into a corridor lined with glass cases, walked back and forth, and talked. The cases contained plaster surfaces, shapes made from metal and taut thread, and mechanical devices fitted with handles. After the break they returned to the classroom. Later, many would meet again in the well-stocked reading room.

A recollection written decades afterward remains a recollection. The survival of the building and models corroborates the setting, not every remembered detail of a particular afternoon. Yet the scene directs our attention toward things that were deliberately arranged. Someone made room in the timetable for an interval. Someone placed mathematical objects where students would encounter them. Someone provided shared books and a place to sit near other people doing the same work.

Conversation during a break was neither commanded nor guaranteed. The arrangements made it easier. This chapter concerns that layer of mathematical history: conditions under which an unfinished thought can meet another person before it becomes a paper, and conditions under which that encounter can leave something for the next generation.

Göttingen's famous past did not ensure such a future. Gauss had worked there for decades; Riemann studied and taught there and delivered his lecture on the foundations of geometry. Nevertheless, enrollment in mathematics and the natural sciences reportedly fell from roughly 240 in 1882 to around ninety a decade later. These are combined subject totals, not a headcount for mathematics alone. The decline mattered even at a university whose name already carried extraordinary prestige.

A reputation does not renew itself. Between Klein's arrival in 1886 and the early 1930s, a collection of practices and resources came together in this small town. There was no blueprint in 1886 specifying the next forty years. Chairs, courses, industrial relationships, libraries, editorial connections, and eventually a new building accumulated through separate decisions. Their coherence becomes easier to see afterward.

Earlier chapters often followed an individual and a manuscript: Archimedes's letters, algebraic challenges between Italian cities, Newton's unpublished calculations, Galois's difficult transmission history. Those stories should not imply that earlier mathematics had no communities, schools, academies, or rapid conversations. Nor was Galois wholly unread during his life. The contrast is one of institutional density, not the sudden invention of collaboration.

What deserves a separate chapter is the way research, teaching, preservation, recruitment, and publication could reinforce one another. In Göttingen, more unfinished work acquired a regular intermediate destination. It need not remain private until it was ready for a journal. To understand that development, the subject of the sentence has to become larger than the mathematician at a desk.

2. Resources before reputations

Felix Klein came to Göttingen in 1886. A common account of his achievement says that he recruited excellent mathematicians. That is true, but it leaves out the most laborious conditions of recruitment. A person needs a position; a position needs funds; funds and appointments had to pass through Prussian educational administration. Klein worked on what the university could make available to mathematics, not only on whom he would like to invite.

He developed advanced courses and seminars and secured shared reading facilities. He also connected mathematical education with concerns beyond the university. After visiting the United States in 1893, he reported to Prussian authorities on higher education and advocated changes involving secondary-school teacher training, access for women, stronger applied mathematics, and support from engineering and industry. These were not peripheral distractions from making a mathematical center. They helped determine whether such a center could sustain itself.

Friedrich Althoff, a powerful figure in the Prussian educational administration, was an important contact. Klein's letter of June 2, 1897, discussed examination regulations and a qualification through which applied mathematics could become part of a teacher's recognized professional preparation. The connection between mathematics and the state passed through such rules as well as through the appointment of famous professors.

In February 1898 an association for supporting applied physics was established at Göttingen. Its contemporary anniversary account gives February 26; later summaries do not always agree in their dating. Mathematics was added to its explicit remit and name around the turn of 1900–1901. The sequence is more important here than forcing inconsistent summaries into a falsely exact date: the organization began with applied physics and then formally widened to mathematics.

The association linked university professors with industrialists and engineers. The chemical industrialist Henry Theodor von Böttinger chaired it, Klein served as vice-chair, and Althoff belonged to the administrative network through which ambitions could become appointments and facilities. At one stage the association connected with about fifty financially substantial industrial representatives.

Three layers of support eventually operated together. State money sustained regular chairs, posts, and buildings. Industrial contributions helped new activities in applied and technical physics and mathematics begin. American philanthropic funding later supported major construction in the 1920s. These sources were not interchangeable, and their arrival at different times should not be compressed into one act of patronage.

Industrial support does not by itself mean that industry purchased control of a mathematics department. In an account by one of the founders, Rieppel, improved training for secondary-school teachers stood prominently among the association's purposes, alongside applied research and responsiveness to practical needs. Better school teaching would prepare stronger university entrants; stronger entrants could feed advanced training and research. By 1910, 178 Prussian students had reportedly obtained the new applied-mathematics teaching qualification.

The resulting cycle connected school education, university teaching, seminars, appointments, applications, and further support. Publication and international contacts added reputation, which could assist recruitment in turn. This is an interpretation of mutually reinforcing arrangements, not a claim that every benefit flowed automatically from a consciously engineered loop. Between 1898 and 1908, the number of regular professorial positions in mathematics and physics grew from roughly five to roughly ten. Again the count combines subjects.

Hilbert's arrival in 1895 also required more than an invitation. Klein had wanted him after Schwarz left for Berlin in 1892, but Heinrich Weber received the position. When Weber moved to Strasbourg in 1895, Klein tried again successfully. An apparently straightforward appointment had a three-year institutional prehistory.

Klein and Hilbert developed different centers of gravity. Klein devoted great energy to organization, education, appointments, applications, and relations with public authorities. Hilbert's force lay especially in research and the organization of significant problems around which younger mathematicians could work. But one was not simply an administrator and the other a thinker. Both researched, taught, edited, and influenced appointments. Their contributions overlapped within the system they helped enlarge.

3. A room whose books stayed inside

The reading room was not just a small branch library. Books and journals were kept available for use on site rather than lent out. Students could spend long periods there, and the room functioned as a shared workplace for mathematics. Klein had used a related arrangement in Leipzig and continued it in Göttingen. A restriction on borrowing served an enabling purpose: the needed volume was more likely to be there when another reader came looking.

Alongside books and periodicals stood material less readily available elsewhere. Courses were often written up by assistants or students, revised, and circulated locally. A lecture series did not necessarily vanish when its semester ended. It could survive as a manuscript account of a subject still being developed, available to students who had never heard the original lectures.

The same room therefore supported several activities. A student could consult a journal, read an earlier course, prepare a seminar presentation, and meet an assistant or fellow student. None of these activities individually was new. Their proximity reduced the practical delay between an idea's first presentation and its availability to another person.

The material fragility explored earlier in this series reappears here in institutional form. Archimedes's text survived through a palimpsest; Newton's private papers waited for later readers; Weierstrass's lectures circulated in student copies. Göttingen did not invent unpublished mathematical writing. It provided more systematic ways of collecting and reusing it. Shelving a manuscript can be an intellectual act when the alternative is that nobody knows where to find it.

The corridor models provided another form of shared material. More than five hundred historical objects survive in the university's collection, including plaster, wood, cardboard, thread, metal, drawing instruments, calculators, and mechanical devices. The oldest cardboard polyhedra date to around 1780, and a large proportion of the collection comes from 1870–1920. The university's collection guide describes their range and materials.

These were working objects, not originally museum decorations. A surface drawn on a board offers one projection; a model can be turned, compared with another, or constructed by a student. The nineteenth-century enthusiasm for mathematical models, promoted by figures including Brill and Klein and supplied through commercial makers such as Schilling, served teaching, investigation, and practical training at once.

A surviving model cannot tell us every discussion it provoked. It does tell us that geometrical thinking had tools and places, and that those tools could be put within reach during ordinary movement through a building. Mac Lane's later recollection catches them while they were still part of daily mathematical life rather than only objects of retrospective admiration.

Applications also occupied physical space. Carl Runge came to Göttingen's applied mathematics chair in 1904; Ludwig Prandtl arrived in the same period and developed work in mechanics and fluid dynamics. Industrial interests and funding helped make contact among mathematicians, physicists, and engineers more frequent. The practical consequence was not simply an additional entry on an examination syllabus. It was a different pattern of neighboring expertise.

Theoretical physics was not swallowed by applied mathematics. Göttingen retained its own traditions, chairs, and institutes in mathematical and theoretical physics. By the time of Born's institute and Mac Lane's visit, proximity and personal overlap could be strong without administrative merger. Institutions stood nearby; people moved among them; problems crossed boundaries; the formal posts remained distinct.

This distinction prevents an easy institutional myth. Shared space is not the same as one organization, and interdisciplinary contact is not the absence of disciplinary differences. The strength lay partly in making those differences accessible to one another. The reading room and the corridor did not dissolve specialization; they provided occasions on which it could speak across its ordinary divisions.

4. The student at the front

Three kinds of gathering must be distinguished. In a lecture, a professor presents and students take notes. In a seminar, participants prepare material and present it for questioning. In a colloquium or mathematical society meeting, visiting or local speakers address an assembled audience. Göttingen used all three, but the second is central to the training mechanism described here.

Klein's seminars left an unusually substantial record: twenty-nine manuscript volumes comprising about 8,600 pages, now digitized. The collection covers decades and includes material from his successive institutional settings; it should not all be counted as Göttingen activity. It records speakers, subjects, dates, and expositions, making it possible to investigate particular meetings rather than rely only on affectionate recollections. The Clay Mathematics Institute reports the collection and its digitization.

Published articles usually display an argument in its finished public form. Seminar records can show its route into discussion: the chosen topic, the order of exposition, an intermediate formulation, or a question that still required work. They do not preserve everything said in the room. Their value is that they retain a structured trace of mathematical labor that a journal article ordinarily removes.

The winter seminar of 1911–1912 examined the historical development of calculus through original sources. Klein introduced the program and distributed topics. The semester produced twenty-three recorded presentations, not necessarily twenty-three separate speakers. Students and younger researchers, including women, participated. History here was itself a seminar activity, requiring participants to reconstruct arguments from documents rather than merely memorize a canonical succession of names.

When Klein became ill, the seminar continued under Schimmack. That small fact exposes an important institutional achievement. Topics had been assigned, participants had roles, and recording practices were established. The meeting no longer depended entirely on its founder being present each week. A person could withdraw without the activity immediately ending.

The novelty was not that students had never discussed mathematics before. Four features were connected: students produced an exposition; peers and teachers questioned it; a written account remained; and participation belonged to a path running through advanced courses, doctoral work, habilitation, and appointments. The seminar was both a site of research communication and a way of learning how to become a researcher.

Nor did Göttingen invent the German university seminar. Berlin and other centers had strong traditions of their own. Göttingen's distinction was the particularly effective combination of seminars with reading facilities, multiple chairs, visiting scholars, doctoral training, and publication networks. Claiming that no other place combined such elements would exceed the evidence. The comparative question is how densely and successfully the components were linked.

This training demanded a change in the student's role. A participant had to decide what an audience needed to know, make the argument explicit, survive questions, and turn the presentation into a record someone else could use. Those skills are not reducible to solving an assigned exercise correctly. They help explain why the seminar could connect classroom work with the production of new mathematics.

Its openness was also conditional. To benefit from being questioned by a room of mathematicians, one first had to gain admission to the room and possess the preparation and financial support to remain there. The same institution that made collaboration effective could give insiders advantages that were difficult for outsiders to obtain. We will return to the gatekeepers after examining another table at which many of the same people sat.

5. The editorial table

Around 1900, a finished mathematical paper normally sought a journal. Mathematische Annalen had been founded by Clebsch and Neumann in 1868, well before Klein arrived in Göttingen. It was not the university's house journal. Its authors and readership were international, which made influence over its editorial decisions more important than control of a local publication.

Klein entered its editorial structure in the 1870s and remained associated with it until 1924. Hilbert joined in 1903; Otto Blumenthal in 1906. The overlap between a university network and a major journal did not turn every published contribution into a Göttingen product. It did mean that mathematicians influential in training and appointments also helped decide which work international readers would encounter.

Editorial practice was less standardized than many present-day systems. Papers might be solicited, sent directly to a particular editor, read by that editor, or forwarded to a trusted colleague for assessment. External reviewing already existed. What often differed was its organization through personal correspondence and professional networks rather than a uniform submission platform. Modern mathematics publishing itself is not uniformly double-blind, so anonymity should not be used as a universal contrast.

An editor could encourage a submission, select a reader, shape the transmission of criticism, and promote acceptance or rejection. Soliciting work and assembling related contributions could also direct attention toward particular problems. These powers did not determine mathematical truth, but they influenced the visibility through which a claim could be checked, used, or made into a research program.

An editorial seat was itself a form of recognition. Yet the occupant did not possess unlimited personal sovereignty. There were fellow editors, publishers, inherited expectations, and reputational consequences. The 1928 removal of Brouwer from the Annalen editorial board shows both the extent and the limits of power. Hilbert could secure a major institutional change with sufficient support, but protests from other editors showed that it was not an ordinary uncontested prerogative.

The intellectual and political dimensions of that conflict deserve their own chapter. Here its relevance is structural. A powerful editor could sometimes act on the composition of the editorial body itself, while collective resistance exposed the contested rules of the institution. Calling the journal simply Hilbert's instrument would erase precisely what made the episode a crisis.

Put the seminar and the editorial table beside the appointments table. Some Göttingen mathematicians helped decide who would be trained, who would obtain a position, and whose papers would circulate internationally. No single decision explains a research center's influence. The overlap of these roles helps explain how that influence could reproduce itself.

It also complicates the language of merit. A strong paper still needed readers and a channel; a gifted student needed access and sponsorship; a research program needed people willing to teach and review it. Identifying these conditions does not explain away mathematical achievement. It explains how achievements became available to a community large enough to build on them.

6. Twenty-three numbered invitations

On August 8, 1900, Hilbert addressed the Second International Congress of Mathematicians in Paris on mathematical problems. The familiar scene in which he recites all twenty-three problems and assigns the twentieth century its work did not occur. He discussed ten in the available time. The complete list owed its enduring life to publication and translation.

The German text appeared in the Göttingen scientific society's proceedings in 1900 and again in 1901. A French translation entered the congress proceedings in 1902. That year an English version appeared in the Bulletin of the American Mathematical Society, translated by Mary Winston Newson. The AMS preserves the translation and its publication note.

Winston had encountered Klein at the Chicago mathematical congress in 1893 and, encouraged by him, studied at Göttingen and earned a doctorate. Her role makes the network visible in the material transmission of the list. The celebrated address reached a new audience through the work of a mathematician whose own training had traveled along the same international route.

Numbering was consequential. The eighth problem or the tenth problem could become a shared reference across languages and institutions. A mathematician in Moscow and another in Chicago could identify the same cluster of questions without reproducing the entire address. Print turned one person's selection into a durable, portable agenda.

It was not an administrative command. Hilbert could not compel the world's mathematicians to work on his list, and many important developments followed other directions. His authority certified the questions as worth attention. The certification drew on a network of research prestige, doctoral training, editorial influence, and international congresses rather than on a single speech's rhetorical force.

American study in Germany grew during the late 1880s and 1890s; Klein's Chicago visit strengthened particular connections. Exact enrollment claims need caution. Formal matriculation, permission to attend lectures, appearance on a professor's course list, and eventual receipt of a doctorate count different populations. Germany is not synonymous with Göttingen, and visitors should not silently become registered degree students.

The more consequential pattern is qualitative. Many visitors did not simply attend one famous lecture and leave. They entered seminars, used reading rooms, prepared dissertations, and formed connections with editors and teachers. Returning home, some adapted elements of this organization to American universities. What traveled included a conception of how mathematical research and training could be arranged, not only a collection of theorems.

Comparison prevents the city from becoming an isolated miracle. Berlin had chairs, seminars, the Prussian Academy, and strong circles around Weierstrass and Kronecker. Its strengths were real, even though the particular connection among mathematics, applied physics, industrial support, and centralized facilities differed from Göttingen's. Difference in organization does not mean a simple opposition between modern and obsolete.

Paris distributed mathematical power among the Sorbonne, École normale supérieure, École polytechnique, Collège de France, and the Academy. The dispersion was part of its institutional form. The Institut Henri Poincaré, established with philanthropic support in 1928, provided a more concentrated center for international contact and mathematical physics. Cambridge, meanwhile, combined colleges, the Mathematical Tripos, and private coaching, with an undergraduate culture long shaped by competitive examinations.

Cambridge research was developing, but the German sequence of doctorate, habilitation, and chair was not its organizing path; the PhD arrived only after the First World War. Both systems were deeply institutional, while rewarding different performances and routes. Rockefeller support for both Göttingen and Paris in the 1920s also shows that concentrating research infrastructure was becoming an international philanthropic strategy, not a uniquely German recipe.

7. The gates

By 1900, research mathematics possessed many features of a profession: recognized qualifications, sequences of positions, state salaries, specialist journals, seminars training successors, national societies, and international congresses. It was not yet a large independent occupation detached from education. Many people trained in mathematics became secondary-school teachers. The research community operated inside a larger system of schools, universities, examinations, and public administration.

What developed was not the first possibility of earning a living through mathematics. It was a more reproducible path by which a trained researcher could help train another. Göttingen made that path unusually dense, but the doctorate, habilitation, assistantship, and chair belonged to a wider German university order. Nor did every stage carry a salary.

A full professor holding a chair normally received a state salary and pension rights, together with substantial influence over appointments, examinations, and resources. Other professorial ranks varied: some posts were salaried, while some titles did not entail a regular civil-service position. A Privatdozent had authorization to teach independently but no salary guaranteed by the title. Fees, family resources, grants, and outside work could be essential.

Assistants, by contrast, could be paid while occupying temporary, dependent positions. Their work and renewal were controlled by senior figures. The ladder therefore did not rise simply from low salary to high salary. A career could pass through years without secure income before reaching a permanent chair. Talent mattered, but the ability to survive those years also depended on resources and patronage.

Gender created a more explicit barrier. In 1874 Göttingen awarded Sofya Kovalevskaya a doctorate on the basis of work completed after private study with Weierstrass in Berlin, where she had been excluded from ordinary university attendance. The oral examination was waived, and she did not reside as a student at Göttingen. The award recognized her mathematics; it did not establish an ordinary route through the institution for women.

Grace Chisholm's 1895 doctorate represented a different kind of entry. She had come to Göttingen in 1893, attended under special permission, studied with Klein, completed research, and sat an oral examination that included mathematics, physics, and astronomy. Mary Winston and another American woman were among the women arriving in that period. Claims about the first woman doctorate must therefore specify discipline, institution, and route. Chisholm was not the first woman in every field to receive any doctorate in Germany.

Even entry as a student did not open the teaching profession. Prussia permitted ordinary female matriculation from 1908, while regulations still barred women from habilitation. In 1915 Klein and Hilbert invited Emmy Noether to Göttingen. She applied for the qualification that July. The obstacle was not an absence of mathematical accomplishment.

Supporters in the mathematical and scientific division sought an exception for her, rather than immediate repeal of the general prohibition. The application went to the Prussian ministry in November 1915; the refusal belongs to the subsequent administrative process, documented in 1917. The institutional record also contains dissent within the wider faculty, so a unanimous mathematics-and-science bloc should not be assumed from the existence of the majority petition. The university's archival inventory lists the application, votes, refusal, and later appointments.

The objections involved more than a technical rule. Habilitation conferred an institutional standing, and opponents resisted women entering that standing. Arguments invoked returning soldiers being taught by a woman and the consequences of admitting women into the professorial community. The supposedly neutral qualification route embodied a prior decision about who could qualify.

When the exception failed, Hilbert made teaching possible under his own name. A course in the winter semester of 1916–1917 was listed as a mathematical physics seminar under Professor Hilbert with Dr Noether's assistance, meeting on Monday afternoons without fees. Formal authorization and actual intellectual work were separated. The workaround depended on a recognized man's position because the ordinary route remained closed.

Hilbert's supposed quip that a university was not a bathing establishment is widely repeated in several versions. His forceful opposition to discrimination is well supported; the exact words are not preserved in a contemporary 1915 minute. The memorable line should not displace the rejected administrative petition, which identifies an actual gate and the rule used to close it.

Noether finally habilitated in 1919. A professorial title followed in 1922 without a regular civil-service chair or its salary. In 1923 she obtained a modest paid teaching assignment that required renewal. Eight years separate the first application from that payment. The university's own account confirms the distinction between academic recognition and a small fixed salary. Göttingen's Noether biography summarizes the sequence.

Economic barriers and recommendation networks operated alongside gender exclusion. Their effects need not be identical to be real. Nor should antisemitic discrimination under the empire be conflated with the legally enforced mass removals after 1933. Earlier prejudice and appointment obstacles coexisted with prominent Jewish and Jewish-background scholars in Göttingen mathematics and physics. The later destruction was a distinct development, not merely a different name for an unchanged institutional condition.

8. The building came last

Visitors today can be shown the mathematical institute building. It came into use in 1929, after many of the practices described here were already decades old. Seminars, reading facilities, prepared lecture manuscripts, protocols, models, chairs, industrial support, editorial networks, and international students preceded it. The most visible object was not the origin of the system.

Courant helped organize the institute in the mid-1920s; planning, funding, and construction between 1927 and 1929 brought previously dispersed facilities together. The organizational step and the building's completion are different dates. Rockefeller-linked funding supplied a major share of the support. The often-cited figure of about $350,000 should be treated as the scale of the philanthropic commitment, not automatically as a verified total construction cost.

By late 1929 the building was in use. Two years later Mac Lane walked its corridors. Reversing that order encourages a misleading picture of a golden age: a collection of geniuses happens to settle in a charming town, talks during walks, and produces modern mathematics. The scene omits the assistants preparing lectures, students keeping records, officials approving chairs, publishers printing, and industrial representatives raising money.

It also suppresses disagreement inside the institution. Klein's efforts to connect mathematics with technology and industry attracted opposition from those worried that pure research would be subordinated to technical-school priorities or practical demands. The tensions were managed through people and negotiations rather than permanently dissolved by one perfect organizational form.

The catastrophe of 1933 later encouraged memory to imagine an intact lost world before it. Sudden destruction can make earlier conflicts and exclusions fade from view. Acknowledging those tensions does not diminish what was destroyed. It makes the achievement and the destruction more accurately describable.

Some measures of activity survive: the expansion of chairs, the applied teaching qualifications, twenty-nine volumes of seminar records, twenty-three presentations in a particular semester, hundreds of models, and roughly seventy-six doctorates commonly associated with Hilbert's Göttingen supervision. These numbers measure aspects of density and training. They do not add up to an objective ranking of mathematical quality.

Papers, books, courses, dissertations, visiting collaborations, and editorial influence have no single common unit. An international journal's excellent contributions cannot all be credited to the university where several editors worked. The argument for Göttingen's significance does not need that inflated bookkeeping. It rests on identifiable mechanisms and their documented effects.

Compare those mechanisms with earlier settings without turning the comparison into a tale of absolute novelty. Scribal schools trained through copying and direct instruction. Correspondence networks connected distant investigators but depended heavily on particular correspondents. Journals carried finished results further than a private letter. Nineteenth-century seminars and lecture-note networks already gave unfinished work some public life. Göttingen linked these functions with unusual intensity.

An idea could be spoken in a seminar, entered in a protocol, placed in a reading room, encountered by a later student, developed into a dissertation, and eventually published. Each transition could fail, but a recognizable route existed. Research did not have to restart from an individual's private papers every time a cohort changed. The chapter's central claim is about this route, not the manufacture of genius.

Copying the final visible products—a building, famous names, prizes—would not reproduce the route by itself. The less photogenic arrangements mattered: a nonlending library, a notebook, a scheduled interval, money for teacher preparation, and a career structure. Yet that structure also excluded. The same authorization that gave an insider a clear path allowed a gatekeeper to deny it to someone else.

The relation is not a universal law that every tighter connection necessarily means harsher exclusion. It is a historical question to be answered at each gate: who controlled it, which rule applied, and who lacked the resources or recognized identity needed to pass? Noether's rejected application provides one unusually clear answer. A useful institutional history must be able to name both the people connected and the people kept out.

This machinery could shape who counted professionally as a mathematician, which problems acquired standing, who was trained, and whose work became visible. It could not decide by organization alone what ought to count as a proof. The community it sustained would soon argue over foundations more fiercely than ever. In the corridors and reading rooms that made collective mathematics possible, agreement about mathematics itself was still to be won.