I measure dielectric permitivvityversus T in bulk BaTiO3 samples .It is known that when it undegoes from paraelectric to ferroelectric(cubic to tetragonal) it exhibits a first order transition and thermal hysterisis is expected. I also observed thermal hysteris on the transitions to lower temperatures, i.e when it undergoes from tetragonal to orthormobic, and orthorombic to romboedric, but aren't these transitions regarded to be as second order and thus, I should not see thermal hysterisis?