See implementation notes. What I evaluate is the sorts of questions and answers the students generate both on their own and in the discussion.
I've been very impressed by the level of discussion this generates. The students love the paper for its readability, the "outdated" ideas, and the cleverness of the hypotheses put forth. They are very good at putting themselves in Lewis' position, knowing what he did, and the discussion quickly developed (this year) to a surprisingly deep discussion of the nature of scientific truth, the iterative scientific method, and the role of theory and experiment. The students "get" these models, and can see how powerful they are even though they are "wrong". I finished by emailing them Bronowski's incredibly powerful monologue on scientific knowledge:
This is G. N. Lewis' classic paper explaining his "octet rule" and the idea of bonds being represented as shared electron pairs. It's beautifully written, and is a lot of fun for students to read. Highlights include his description of atoms as being concentric "cubes" of eight electrons at the vertecies, philosophical discussions of the importance of letting experimental observations guide the development of theory, and the sense that students gain that Lewis developed this theory completely in the absence of an orbital or Bohr-type model of the atom. It wonderfully captures the way in which a brilliant mind wrestled with the problems of developing a bonding theory for the periodic table. Students understand the paper pretty easily, and are capable of picking out little "gems" on a first read of their own which they bring to a discussion.
Students should be able to address the following questions:
(1) Explain Lewis' original "cube model" and his subsequent "tetrahedron model" of the atom.
(2) Why does Lewis abandon his first for his second model in this paper, even though it contradicts the prevailing theory of the day.
(3) Compare and contrast what Lewis actually proposed with the "Lewis Dot Model" which youlearned in genchem.
(4) Please discuss the usefulness of low-level theory which is easy to use vs. high-level theory which is hard to use, and articulate the value (or lack thereof) of doing things at a "simplistic" level.
I hand out this paper in advance of one class and have the students bring the answers to the questions in the handout as well at three questions of their own, which they give to another student. By the next class, they try to answer those student questions, and we use all of the questions so generated as a basis for a full class discussion.