Comment on ''Minimal Surfaces, Screw Dislocations, and Twist Grain Boundaries''

In a recent letter, (Phys. Rev. Lett. 82, 2892(1999); cond-mat/9808306) Kamien and Lubensky calculated the energy of the surface constructed via a linear superposition of screw dislocations in SmA phase, and obtained the positive (repulsive) sign of the long-range interaction between dislocations. We would like to object that this interaction is attractive. This conclusion can be important to explain the discontinuity of the TGB - SmA phase transition in chiral liquid crystals.