The relationship of cell organelles to valve morphogenesis was investigated in three species of Nitzschia. One, N. sigmoidea (Nitzsch) W. Sm., showed consistent ability to generate both nitzschioid and hantzschioid symmetry in daughter cells following cytokinesis; the other two maintained nitzschioid symmetry stably. From previous work with Hantzschia, a certain sequence of events could be anticipated in the cytoplasm. In two significant areas–the behavior of the Microtubule Center (MC) and its microtubule (MT) system, and the central origin of the silicalemma–not only were the results unexpected, but the three species showed fundamental differences among themselves. In N. sigmoidea, the silicalemma (and the future raphe region) arises centrally on the cleavage furrow, and after some lateral expansion, the silicalemmas and their associated organelles move in opposite directions in daughter cells, so that the raphe and the raphe canals end up along the girdle side of the cell as expected. However, the MCs never become associated with their silicalemma, remaining throughout near the girdle bands. In N. sigma (Kütz) W. Sm., the silicalemmas arise centrally and after lateral growth, move in opposite directions to generate nitzschioid symmetry. In this case, the MCs move to the vicinity of but never close to the silicalemmas, and follow them distantly during their lateral movement. In N. tryblionella Hantzsch, the new silicalemmas arise opposite one another, on one side of the daughter cells; each MC soon moves very close to its silicalemma, and remains thus through most of valve morphogenesis. Later, only one silicalemma/MC complex moves laterally, establishing the nitzschioid symmetry in both daughter cells. In all three species, as in Hantzschia, linear arrays of mitochondria aligned along MTs occupy the forming raphe canal, and microfilaments line the outer edge of the expanding silicalemma. The fibulae (the wall struts arching across the raphe canal) in Hantzschia always grow from the valve surface to the girdle surface of the forming valves. In these three Nitzschiae, this invariably happens in only one daughter cell of any pair; in the other, all the fibulae grow from the girdle surface to the valve surface. An explanation of these variations is proposed: that the morphogenetic machinery of Nitzschia and Hantzschia have a common origin, with present Nitzschiae having undergone considerable diversification at the intracellular level, causing the unstable cell symmetry exhibited by several modern species. Perhaps a taxonomic distinction between Hantzschia and Nitzschia lies in whether the morphogenetic machinery associated with valve morphogenesis moves laterally in the same or in opposite directions.