It is funny how he appears, judging from his diagrams, to use "x+y" to mean AND, and "xy" to mean OR. This is a wholly self-consistent definition compatible with boolean algebra, but is the opposite to the convention since adopted.