Summary: Hyperbolic polynomials elegantly encode a rich class of convex cones that includes polyhedral and spectrahedral cones. Hyperbolic polynomials are closed under taking polars and the corresponding cones-the derivative cones-yield relaxations for the associated optimization problem and exhibit interesting facial properties. While it is unknown if every hyperbolicity cone is a section of the positive semidefinite cone, it is natural to ask whether spectrahedral cones are closed under taking derivative cones. In this note we give an affirmative answer for polyhedral cones by exhibiting an explicit spectrahedral representation for the first derivative cone. We also prove that higher polars do not have a determinantal representation which shows that the problem for general spectrahedral cones is considerably more difficult.