You could perhaps base it on articles by Martin Gardner and others. I taught a non-honors course where the first half was group theory and then the second half was evenly divided between Flatland and probability. The group theory section used Gardner's "Ambidextrous Universe" for readings, the second section used Flatland, of course, and the third section used Keith Devlin's book "The unfinished game". All three sections were augmented with articles by Gardner. You can find my lecture notes and other course materials on my website. http://www.colby.edu/personal/s/sataylor/teaching/S10/MA111/