Feynman's missing method for third-orders?

I was reading about Hans Bethe in an old but good book the other day and was excited to learn that one of the first things Feynman did at Los Alamos was develop a numerical method for solving 3rd-order ordinary differential equations. This is also mentioned in other works here and here. I'm sure the method isn't anything special by today's standards, but it seems like it might make a good story/lesson for a class. Unfortunately, none of the references on-line seem to explain exactly what the method was or any more of the mathematics. Anybody know about this and able to fill in the blanks or point to where they are filled in?