Probable reasons why some protein crystals grown in a magnet exhibited better quality than control are discussed as follows. (1) Sedimenting three-dimensional nuclei are able to have the same orientation as the underlying, mother crystal into which the nuclei merge. (2) Protein solution may become more viscous, leading to reduction of convection. (3) If an upward force is generated by use of an inhomogeneous magnetic field, the effects of the density differences can be made less significant, causing the reduction of natural convection and the retardation of crystal sedimentation.