He hated engineering. At that time, he studied engineering because there was no mathematics course. But he also had a family issue. In my grandfather's mind, he had an uncle called Cândido, who was an engineer. Since my father was Cândido and he should have been an engineer too.

... as a child, the family influence was much more French than anything else, so everyone spoke French.

... at that time the pressure from the Catholic Church was very strong, and my parents feared that in primary school I would be discriminated against for not having the current religion.

We had two wonderful ones! One, which was my mother's, was an upright Steinway, magnificent! And another, which had been an aunt's, my mother's sister, was an old quarter grand Erard. They were two wonderful pianos. The Erard stayed, but I learned to play the other. My mother taught me and my sister. They were remarkable: the two pianos ... and the books!

He was much stricter with me than with anyone else. But he was an excellent teacher; he was a famous teacher. He was stimulating and very demanding: he gave much more content than usual. I studied the theory of limits with him in my fifth year of high school, which was later taught in my freshman year of college. He gave my class lessons on limits, derivatives and was very keen on proofs, geometry ... And he understood it! He had studied in Europe, he knew more than most.

Elza became interested in Physics, influenced by the success and popularity of that science ...

I was lucky enough to be guided by a Bourbaki, Jean Delsarte. Delsarte was a lesser-known Bourbaki, but of those who were here he was - perhaps - the one who was most concerned with the transmission of knowledge. Something that Weil, for example, was less good at. If we took advantage of his science that was very well, but if we didn't, so much the worse; it wasn't his concern ...

This conjecture has been demonstrated for curves and for other varieties of a particular type. (A Weil, 'Les Courbes Algébriques et les variétés qui s'en déduisent' Ⓣ, and 'Numbers of Solutions of Equations in Finite Fields'). This work contains research on this conjecture.

They don't have entrance exams there. Passing the baccalaureate, one enters the university. The university's mathematics course does not train its future teachers. Almost no mathematician has taken the university course. Next to the university, they have the elite institutions, which are the 'grande écoles'. These have a very tough entrance examination, which actually starts from high school. Those who are very good, already start in special classes and end up being able to go to the École National de Administration, from where presidents, ministers, etc. come out; to the École Polytechnique, which produced the top people before they started to be produced by the École Normale Supérieur, which trains researchers in mathematics and physics. I believe the French university system cannot be copied; for it is a very peculiar model.

It didn't take long for me to start teaching a class. It so happens that Catunda went to the United States for an internship and during his absence Lyra and I took theoretical classes. That was maybe in the early 50's [Note by EFR: actually it was 1946-47]. I mean that after four or five years I already took classes, and since then I've continued ... I think it was there, when he went to the United States, that I started to take theoretical classes. That was the common relationship. At first the assistant did exercises and studied; and gradually, over time, they would take over the classes as well. ... I started teaching calculus exercises - Catunda gave the theoretical part, and I taught the exercises as his assistant. Sometimes I attended his class; he would then prepare the exercises I was going to do. I think it was four exercise classes a week. There were a lot of classes, four theoretical and four exercises... something like that.

The course started with the construction of real numbers. This is a beautiful construction, but extremely elaborate. I remember very well that I thought the idea of Dedekind cuts was wonderful, but making the students' first contact with calculus by defining operations and their properties based on the notion of cuts is a tragedy! ... It was over the students' heads. ... The course only began to be of interest when it entered the notions of derivative and integral.

The Mathematical Analysis course should be given after the Differential and Integral Calculus course; first give a course in Calculus without criticism and then do Analysis, which is actually the justification, the critique of Calculus, with all the theorems, the construction of real numbers, etc. After all, how could students directly understand the criticism of something they didn't even know yet? ... I insisted with Catunda that it was better to start with Calculus than with Analysis and he agreed. He was very liberal; very few full professors gave their assistants the freedom he afforded us.

Her room at the Institute of Mathematics and Statistics at the University of São Paulo was always open, whether she was inside or not. When the place was empty, colleagues simply went in to borrow a book or read the newspapers. Whoever found her there took the opportunity to talk or ask for some of her wise advice. For professors at the Institute of Mathematics and Statistics, that unlocked door symbolised their open and creative thinking. From her election to head of the Mathematics Department in 1968, during the military regime, she began to act more in matters related to teaching. She had begun to take an interest in the question when she realised that modern mathematics, in her own words, "was doing a lot of damage." She became very involved when the Ministry of Education decided to impose the Licentiate in Science, which she thought would be extremely harmful, especially to Mathematics. This involvement, added to the problems caused by the political situation, with threats from the military regime on the one hand and the pressure of student ferment on the other, made her no longer available for research activities. According to her, at the time it was already very difficult to "keep your nose above water". The separation of the Bachelor and Licentiate courses in Mathematics was finally effected in the late 1960s.

... this is not just any job, we start at nine in the morning and finish at five in the afternoon. Either you are engaged, absorbed in work, or you are not. If you are, it sometimes damages other relationships, and that's the kind of accusation you get. I think everyone who is in this life has that experience. Those who asked me were more distant relatives and friends who wanted to meet me and I didn't have time. ... I realised that it was difficult to maintain certain relationships and at the same time work as I wanted. This is difficult.

I now have other ideas about the best way to go about things. I was part of a tradition - which I thought was very good - of the Analysis of the Italian school. Now I like to think more about applications, about history. ... Not that I think it's essential to tell a story, it's not that much that is required. It's nice for us to tell a little about the history of the subject, but now that I know a little more about the history of mathematics, it serves as an instrument for reflection. You can see what is the best way to open up the subject to students by thinking about the way things developed. So today I would follow other paths, but I don't think the other ways were wrong.