Mathematics at Duke: The 1940's

In just a decade after its establishment in the mid 1920's, Duke had developed a strong national reputation as a selective and expensive university - its tuition was $200 per year with an estimated total costs per year of $600. As noted in Chapter 2, enrollment increased 75% in the decade of the 1930's and attracted many students from the Northeast and even a few from the West. The male-female (Trinity-Women's College) ratio stayed relatively constant at two to one with 1710 men and 866 women undergraduates in 1939.

It was also something of a party school. The May Day celebrations with its beauty contest and the dances and fraternity parties were high points of the school year. Les Brown and his Blue Devils became the hot student jazz band. The grade point average remained very low by current standards with a significant number of undergraduates unable to maintain the requisite C average needed to graduate. Many students needed to take College Algebra, Trigonometry and Analytic Geometry before they were allowed to start in Calculus.

During the 1930's, Wallace Wade and Eddie Cameron had built up the football and basketball teams to national status. The 1938 Iron Dukes, who had not been scored on at all during the regular season, played USC in the Rose Bowl in January 1939, and suffered their only loss in the last minute of play in that game. Three years later, the undefeated Blue Devils were again invited to the Rose Bowl. After Pearl Harbor on December 7, 1941, fearing a possible Japanese attack on Southern California, the game was moved to Durham, the only time it was played outside of Pasadena. Duke lost to Oregon State in a cold drizzle by a score of 20 to 14. [Durden - 226]

After the start of WWII, Duke University joined most other college campuses in modifying their curriculum to assist the war effort. An accelerated three year degree program was instituted for Trinity College with three terms per year and few breaks. For example in 1945-46, the terms were from July 4 (sic) to Oct 23, November 3 to February 23 and March 4 to June 22. Christmas break lasted only from Saturday December 22 until Thursday December 27. Although many Trinity students were drafted or volunteered for the military, this drop in male students was more than compensated by the Navy officer training program that reached 1600 naval trainees by Fall 1943, one of the largest such programs in the nation. At this time, Duke had about 1400 men and 900 women regular undergraduates. Since this and other military related programs required a full year of mathematics, class sizes in that department grew very large. Some instructors were required to teach fifteen to twenty hours per week [Durden 461].

Although the Trinity men met year round, the calendar for the Women's College remained the traditional mid September to mid May with a full two weeks for Christmas break. This made it difficult for women to take many advanced courses in math and science. Few took the bus to West Campus for math courses and very few majored in math. Women did take advantage of the war shortages to demand an end to the requirement that they wear silk stockings in public, much to the dismay of Dean Alice Baldwin. [Durden 455]

For much of the 1930's and 1940's, the mathematics department consisted of four research professors who taught the graduate courses, Leonard Carlitz (Number Theory and Combinatorics), John Gergen (Analysis), John Roberts (Topology) and J.M. Thomas (Differential Equations), and up to a dozen other professors, instructors and visiting faculty who did the bulk of the undergraduate teaching. Carlitz, Gergen and Thomas were strong personalities and did not talk with each other either professionally or socially. Gergen remained chair from shortly after he was hired and ran the department authoritatively, assigning courses, and rarely holding department meetings. Roberts was more congenial.

A thesis was required for a Masters of Arts in Mathematics from the start of the University through much of the 1960's. During the 1930's, the graduate program produced 56 MAs of whom 25 were women. With the disruptions due to the war, the number of advanced degrees decreased significantly with no graduate degrees awarded in 1943. Production picked up in the 10 years after the war with 52 MAs awarded of whom 11 were women. Relatively few students who received a Masters continued on for a Ph.D. Only 9 Ph.D.s were awarded in the 1930's and 20 got their doctorates in the postwar decade.

As noted in Chapter 1, the first math Ph.D. at Duke was awarded to Ruth Stokes in 1931. In 1933, the second and third Ph.D.s were awarded to long-time Duke professor Francis Dressel and to Mabel Griffen. No other female received a Ph.D. in mathematics at Duke until Lois Reid received hers from Carlitz in 1967. During this period from 1931 to 1967, 50 men received doctorates in mathematics at Duke. Although Reid was the only female Ph.D. among the first 39 of Carlitz' students, five of his last six students were female.

The end of the war and the GI bill brought a huge influx of veterans to Duke. In 1946, over 2500 men enrolled, bringing total enrollment in the University to a record 5100 students and eclipsing the prewar high total enrollment of 3685 students in 1940 with three quarters of them undergraduates.

One of those students was a veteran not of the US military, but a refugee from Austria who had served, after a fashion, in the French and British military. After Hitler's rise in Germany and the Anschluss of Austria in 1938, Walter Rudin and his sister were sent to school in Switzerland on a temporary visa and their parents followed soon after. Much to their dismay, the Swiss would not renew their visa. With the help of friends they managed to get permission to move to France shortly before the deadline. When war broke out through Europe, they were forced to move again. In As I Remember It, Rudin chronicles his adventures escaping from the advancing German troops and making his way to England on one of the last ships allowed to leave France. While there, he enlisted in a branch of the British Army for foreign nationals in 1940 and later joined the British Navy where he served as a radio man and translator for several years on ships in the English Channel.

In 1945, Walter made his way to New York where his parents and sister had lived since fleeing from the Nazis in 1941. At that time, his sister was a graduate student at Duke studying Chemistry. After a family reunion, Walter went down to Duke with his sister. He talked with some Duke math professors and then went straight to the Dean to tell him that he knew it all and he wanted to spend a year there to complete the bachelor's degree. Audacious as this request was for one without even a high school diploma, he finally persuaded the Dean to accept him on probation with Junior status. He immediately enrolled for the second (November 1945) term. Four years later, he had not only obtained his Bachelor's, Master's and Doctorate, but was appointed as an Instructor teaching 12 hours a week for the "attractive" salary of $3500.

In 1950, Walter went to MIT as a CLE Moore instructor. When he was assigned to teach a course in analysis, he complained to the department chair that no decent textbook was available. The response: "Why don't you write one" and three years later his classic, Principles of Mathematical Analysis, was published. Toward the end of his two year instructorship at MIT, Walter accepted a position at the University of Rochester.

As an instructor at Duke, Walter was assigned to small cubicle in a former classroom. He got to know and spend a lot of time with another young math instructor in a neighboring cubicle who had been hired in 1949.

Mary Ellen Estill grew up in a tiny Texas town and graduated in the middle of her high school class of 5 students. At age 16, her father took her to Austin to attend the University of Texas. As she relates in More Mathematical People, the registration line was very long for those interested in the liberal arts. Since few were waiting for the elderly gentleman behind the mathematics table, she wound up there in front of Professor R.L. Moore. After long and probing questions, Moore not only signed her up for his courses, but guided her – under his "total control" - throughout her undergraduate and graduate studies. Although she enjoyed the attention at the time, she realized later that, by learning under the "Moore method", she came out knowing virtually no algebra, analysis or even topology other than that in the unconventional Moore language. "I didn't even know the standard definition of compact."

Moore had been chatting with J.M. Thomas on a train ride after a math conference. Duke had no female graduate students in the Ph.D. program and few female math majors at the time and Thomas felt pressure to hire a woman mathematician for the Duke Women's College. This was in sharp contrast to Thomas' stance a decade earlier. With similar pressure to hire a women mathematician after the death of Julia Dale, Thomas wrote to Dean Wannamaker that there was "no woman mathematician who has been actively engaged in research for at least the last ten years and none who gives any promise of becoming such a mathematician." [Durden 101]. Instead, Thomas and others pushed to hire John Gergen in 1936, a decision that Thomas clearly regretted some years later.

However on this train ride in 1949, Moore told Thomas that he had the right person for him and informed Mary Ellen that she would be going to Duke the next fall. She dutifully went, was given a room in the graduate women's dorm, and spent the next four years as an instructor teaching three courses a semester at or below the calculus level. She was quite happy as this was a lighter load than she had been used to as an instructor at Texas while finishing her thesis.

As mentioned above, personal relationships between the senior professors were strained and each had their own research program and graduate students. Although they did not talk with each other, each of the four senior professors was very supportive of Mary Ellen and Walter and the 8 or 10 graduate students. Thomas would occasionally invite the graduate students and junior faculty to play on his back-yard tennis court. John Roberts had several parties at his house each year for graduate students and instructors. He was the only senior professor willing to talk with everyone. Mary Ellen and Walter did get together frequently with several of the graduate students for lunch.

During her years at Duke, Mary Ellen does not remember any visits or colloquia by outside mathematicians. One year, John Roberts ran a topology seminar with Walter, Mary Ellen and a few graduate students. Fortunately, Burton Jones, a former student of R.L. Moore and professor at Texas, had moved to UNC the same year that Mary Ellen joined the Duke faculty. She attended the weekly UNC math colloquia and talked frequently with Jones. She spoke at a couple colloquia at the University of Virginia and attended math society meetings whenever possible where she often had the opportunity to collaborate with Paul Erdos.

While at Duke, Mary Ellen had done ground breaking work related to Souslin's conjecture and was able to find a counterexample to a conjecture by R. L. Wilder at the University of Virginia. Her reputation as an extraordinary problem solver became widely known and, with the help of Wilder and others, she was offered an NSF grant to study at the University of Michigan. However, shortly thereafter Walter had proposed to her. Since Walter had accepted a position at the Rochester, Wilder managed to transfer the appointment from Michigan to Rochester.

With the great influx of students after the war, there was a severe lack of space for the new faculty and the classes were crowded. Several math instructors had cubicles for offices in a building on East Campus even though all math courses were taught on West. With the end of wartime restrictions, the university went on a building spree. A building for the engineering school (Hudson Hall) was begun in 1946 followed soon after by other buildings in the red brick "neo-classical" design on what was to become Science Drive. Construction on the Physics Building was begun in March 1947 and completed in April 1950 at a cost of $948,000. The style of these buildings outraged many alumni with its abandonment of the Tudor Gothic style. However a stone building would have cost twice as much and the desperate need for space and shortage of funds overrode these concerns. These buildings, as was the rule in those days, had no air conditioning. Some professors kept towels in their offices to wipe the sweat off their face and arms.

Classes remained overcrowded and professors began to complain about their low pay in the post-war inflationary period. Tuition doubled and total annual expenses approached $1000. It was necessary to offer some financial aid and alumni responded to requests with generous donations.

The next major task of the math department and many other departments was to recruit many more research faculty for the rapidly expanding undergraduate and graduate programs.

References

Robert Durden, The Launching of Duke University, 1924-1949, Duke University Press 1993.

Walter Rudin, The Way I Remember I, American Math Society 1997.

Interview with Mary Ellen Rudin in More Mathematical People, Harcourt Brace Jovanovich, 1990.

Bulletins of Duke University 1930 - 1955.

Personal communications from Mary Ellen Rudin and Seth Warner were quite helpful.