.. .. iill jS.. TOglr) i
Some manipulation shows that the form which the equations (5.40) take is the same as (5.37) but with a term l/W. added to each
J equation in turn. In other words, they take the same form as the general equations with n = -1, which gives an infinite gradient pick-up at the origin when the growth of W. from an
J initial disequilibrium position is plotted as a function of time. It is possible to give a Lagrangian which generates (5.37) exactly, but this turns out to be not easy to interpret. This point is discussed in Appendix 2.