ABSTRACT

In commenting on an earlier version of this paper, Bob Dorfman remarked to me that I had presented two papers concealed as one. He is, of course, quite right. The second paper begins with section 5.