The Diatom Project, Figure 32.

Fig. 32. Ontogenetic asymmetry of the valve in naviculoid diatoms. One side of the raphe-sternum and hence of the valve can be considered primary, in that it is formed first (a). The secondary side is produced later by a different process, involving the fusion of arms extending out from both centre and poles (b), (c). Where the arms fuse, faults often occur in the stria pattern (compare (c) with (d): these are the Voigt discontinuities. Where both polar raphe endings are turned towards the same side, as in most raphid genera, they turn towards the secondary side ((d). see also Fig. 33). (e)-(j) The evolution of the raphe: (e) the araphid ancestor with rimoportulae at each pole: (f)-(h) differentiation of the raphe from one rimoportula at each pole. Note the finer striae associated with the raphe because of its essentially transapical orientation, the rotation of the raphe towards the centre, and the parallel movement of the undifferentiated rimoportulae to an apical position. (i). (j) Abutting of the raphe onto the strenum and final integration into the primary pattern-centre. In (e)-(j) the pattern-centre is shown stippled and the raphe slits are outlined heavily in black (from Mann, 1984b); primary side at right.