2 HENRY C. WENTE
where (assuming that c o £ 0) a(u), /?(u) will be solutions to a
second order system of differential equations from which the
solution co(u,v) is to be recovered. The system is
a" = aot - 2a 2 /? - 2A/?
( 1 . 4 )
/?" = a/? - 2a/? - 2Ba a = c o n s t a n t .
Furthermore one finds
, . . , , 2
2 , 2 CO . . „ , 2 .
_ . 00 ^ . -00
( 1 . 5 ) 4oo = - (4A + a )e - (4B + /? )e - 4 a ' e + 4/?'e + 6 ^
wher e 6y = 6a/? - 4 a .
If one solves the system (1.4) then (1.3) and (1.5) may be
used to recover oo(u,v). This development is carried out in
Section II. The system (1.4) is an algebraic completely
integrable Hamiltonian system. To solve it we follow the method
described in Darboux  which is to solve the relevant
Hamilton-Jacobi equation by the method of separation of variables.
A general feature of any solution oo(u,v) obtained in this manner
is that it will be periodic in the v-direction (with infinite
period allowed) and quasi-periodic in the u-direction.
In section III we study the case with mean curvature H = 1/2
in R . Here the Gauss Equation (1.2) may be written in the form
(1.6) Aoo + sinhoo coshoo =0, A = B = 1/4.
This is the P.D.E. which initially attracted my attention when
constructing immersed cmc tori in R . One finds here that if
an immersion of Enneper type is defined locally then it extends to
a global mapping of the plane into R . In particular the
immersion will not develop any umbilic points. One can explicitly
compute the centers and radii of the spheres determined by the
lines of curvature from the solutions to the system (1.4). It
turns out that any immersion has an axis I on which the centers
of all spheres lie. This fact enables one to give fairly explicit
formulae for the immersion itself.
All solutions to the elliptic sinh-Gordon equation (1.6)
which are doubly periodic have been classified recently in a