Commentary: The making of Chapter 46 of The Feynman Lectures

In the fall of 1961, the usual two-year course in physics for Caltech undergraduates was replaced by a two-year lecture series by Richard Feynman. The lectures were to be documented for future students. In 1963 a plan emerged to turn the lectures into textbooks, The Feynman Lectures on Physics . Feynman’s responsibility largely ended with the lectures themselves. Each lecture was tape-recorded, and the blackboards photographed. The tape was then transcribed (apart from Feynman’s expletives) into a triple-spaced typescript of the lecture.

Caltech professors Robert Leighton and Matthew Sands were responsible for producing the edited lectures and the final manuscripts for the books (see the article by Sands, Physics Today, April 2005, page 49 ). The initial plan was to assign a particular Feynman lecture to a graduate student, who would attend the lecture, edit the typescript, and supply drafts of the figures based on Feynman’s blackboard diagrams. I was in my sixth year at Caltech when I was asked by Leighton to edit one of the lectures. I don’t know how the grad students were chosen, and I can’t name another who was.

My assignment was to edit “Ratchet and Pawl,” the 46th lecture of Volume I. It is the final of eight lectures related to statistical mechanics and thermodynamics, and in a way it is their culmination: Feynman applied the previously taught statistical mechanics and thermodynamics concepts to dissecting the physics of a device that at first appears to violate the second law of thermodynamics.

Feynman’s ratchet-and-pawl machine is what’s called a perpetual motion machine of the second kind. It conserves energy but converts thermal energy to mechanical energy in violation of the second law of thermodynamics. As can be seen from the figure I produced to go along with the text, the machine consists of a little paddle wheel on an axle along with a notched gear. Random fluctuations in the velocity of air molecules striking the paddlewheel provide torques in one direction or another. A spring-loaded pawl keeps the gear from turning backward. The one-way motion means that work can be extracted to raise an object, such as the bug I drew in the figure.

The extensive markup of the lecture transcript took me a full week. Feynman is subtle! He spoke at many levels at once, with asides, and his style was quite casual and conversational. (One quip: “It turns out, if we build a finite-sized [Maxwell’s] demon, that the demon himself gets so warm that he cannot see very well after a while.”) Leighton provided me with notes containing suggestions as to how to make the material more suitable for publication. The most common change was substituting “we” for his frequent “I.” Contractions were to be eliminated, and phrases like “figure out” replaced with more specific verbs.

A comparison of my markup with Chapter 46 in The Feynman Lectures shows that virtually all my suggestions were adopted. For instance, the transcript on page 46-28 , says, “In this box of gas we look at the molecules and we see, of course, that they are like that.” My version was “Look again at our box of mixed red and black molecules.” The final version was, “Look again at our box of mixed white and black molecules.” I was especially happy to see my rendering of the ratchet-and-pawl machine survive intact, complete with the puzzled bug.

My most serious failing as an editor concerned a set of equations that appear in the published text. In the lecture Feynman says, “As an example, we calculate the entropy difference between a gas at a certain temperature at one volume, and a gas at [the same] temperature at another volume. . . . Well, I will let you do it because I am running out of time.” He then goes on to sketch the equations in words. I noted “do right” on page 46-22 of the transcript, but I never followed through with the equations.

Whereas most of his preceding lectures simply come to an end, Feynman ends this one more philosophically, speculating that the order in the universe we see today is not a statistical fluctuation but a memory of initial conditions. He further explores a concept introduced at the end of Lecture 1-3 , that “the entire universe is in a glass of wine, if we look at it closely enough.” He later continues: “But because we pull up the shades and let the light out, because we cool off on the Earth and get heat from the Sun, the ratchets and pawls that we make can turn one way. This one-wayness is interrelated with the fact that the ratchet is part of the universe. . . . It cannot be completely understood until the mystery of the beginnings of the history of the universe are reduced still further from speculation to scientific understanding.” I was especially pleased to see that, for the most part, those difficult final paragraphs survived my editing .

Further editing of my markup was done to obtain the version finally appearing in The Feynman Lectures. The chapter was broken into sections and equations were added, as was the important Table 46-1, which summarized the operation of the ratchet and pawl. Presumably Leighton did that work, probably with the help of one or more of the people mentioned in the acknowledgments of the published lectures.

I was surprised to find that there is no mention or acknowledgment of me or my fellow volunteered student editors in any of the three volumes of the lectures. When I asked a Caltech archivist about the omission, I was told that the student editors didn’t do well and that the approach was abandoned. Decades later, Sands told me that he and Leighton did the editing themselves. I find it a curious rewrite of history.

Don Groom is a retired senior scientist in the physics division at Lawrence Berkeley National Laboratory. He is indebted to Jeffrey Anderson, Dean Robinson, and Paul Schaffner of LBNL for creating the Feynman document webpage and for dealing with numerous questions and missteps.