The Cayley surface: the equation of the dual surface

This page is dedicated in memory of my friend Sevín Recillas,
who got me interested in this surface.


 
 
 
 
 
 
Scroll down
 
to the
 
discussion
 
of the dual equation.
    Viewing hints:
 
  • This is a large figure and it simply may load slowly. So, the first thing to try is just to be patient for a little while.
     
  • If nothing loads except the light blue background, it may help to (i) click on your browser's "back" button, and then (ii) click on the "forward" button to get back to this page.
     
  • If the picture absolutely won't load, please click on the Static figure link.
  • Implicit equation links: Other Cayley surface links: Other:  

    The equations of the Cayley surface and of the dual surface:
     
        The approach used here is based on what I have learned from Sevín Recillas. 

        We start with a version of the implicit equation of the Cayley surface which is valid at points of the Cayley surface where all four coordinates are nonzero.  The idea is to divide the implict equation by  wxyz,  so that we work with the implicit equation  G(w,x,y,z) = 0,  where  G(w,x,y,z) = 1w + 1x + 1y + 1z.  This leads to the following version of the duality map:
     
          (w:x:y:z) ---> (Gw: Gx: Gy: Gz) = (1w² : 1x² : 1y² : 1z² ).
     
    (A minus sign has been cancelled in the last homogeneous coordinate vector.) 

        Since  1w² = (1w)²  (and similarly for the other variables), our current version of the implicit equation of the Cayley surface leads to the following version of an implicit equation of the dual surface:
              _    _    _    _
             w + √x + √y + √z = 0.

    We can transpose two terms to the opposite side of the equation, and then square both sides:
              _    _     _    _
             w + √x = -√y - √z ;
                   __             __
             w + 2√wx + x = y + 2√yz + z.

    We can isolate the two square roots on one side of the equation, and square both sides again. Doing this and a little bit of simplifying, we obtain the following:
               __     __
              2√wx - 2√yz = -w - x + y + z;
             ____
         - 8√wxyz = w² + x² + y² + z²
                     - 2wx - 2wy - 2wz
                  - 2xy - 2xz - 2yz.

        If we square both sides of this equation, we clearly obtain the following result:
     
          64wxyz = (w² + x² + y² + z² - 2wx - 2wy - 2wz - 2xy - 2xz - 2yz.
     
    This equation is symmetric in the variables w,x,y,z

        Another nice feature is that this equation clearly exhibits the fact that each of the coordinate planes intersects the dual surface in a plane conic curve;  more specifically, each coordinate plane is tangent to the dual surface at every point of the corresponding curve.  In other words, this intersection occurs with multiplicity = 2.  For instance, the equation of the intersection of the dual surface with the plane  w = 0  (obtained by substituting  w = 0  into the implicit equation) is:
     
          (x² + y² + z² - 2xy - 2xz - 2yz = 0.

        On the other hand, using this version of the implicit equation of the dual surface to study the singular locus of the dual surface seems (to me, anyway) to be excessively complicated.  On a separate page I will present a change of coordinates that yields a version of the implicit equation which exhibits the singular locus much more directly. 

     


    The Java files used in this page were downloaded from the Geometry Center webpage.
    I generated the geometric data for this figure in March 2009.
    Latest updates May 15, 2018.

    Prof. Joel Roberts
    School of Mathematics
    University of Minnesota
    Minneapolis, MN 55455
    USA

    Office: 109B Vincent Hall

    e-mail: roberts@math.umn.edu

    http://www.math.umn.edu/~roberts