The frustules of the Diatomaceæ are constantly undergoing division when in vigorous growth. After the frustules have divided, the new ones either separate entirely, as is perhaps most commonly the case; or they remain united, sometimes completely, so as to constitute a filament (Pl. V. fig. 23), while at others the frustules cohere only at the angles (Pl VI. fig. 23), so as to form a zigzag chain. In some species, the frustules are attached to foreign bodies by means of a gelatinous cushion (Pl. V. fig. 5; Pl. VI. fig. 7); while in others they are situated upon a simple or branched gelatinous stalk (Pl. V. fig. 17) or stipes (_stipes_, a stem). When the frustules are examined in the living state, the cell-contents resemble those of ordinary vegetable cells, excepting in regard to the colour, and exhibit granules and globules, and sometimes a nucleus is visible. It will also be noticed that many of the free frustules move slowly across the field of the microscope; but the cause of the motion is unknown. When the frustules have been properly prepared, the surface of the valves exhibits a number of coarser or finer markings, consisting of dots, lines (striæ), flutings, or networks, &c., arranged with great regularity and symmetry, often of extreme minuteness, and rendering them exquisite objects under the microscope. The exhibition of these markings requires not only that the valves shall be properly prepared and mounted, but that the object-glasses be of good quality, and that the management of the light be thoroughly understood; so that to a beginner, their examination is often a matter of great difficulty; for only the very coarsest or largest of these markings can be perceived in the natural frustules. The appearance of these markings, and even their apparent absence or presence, frequently depends upon the kind of illumination used: thus, under one kind of illumination the valves may appear simply white or coloured, while under another they appear covered with lines, and under a third with dots. It will often be observed, also, that the colour of the valves varies according to the illumination and the power used--the same valve appearing white, yellow, brown, blue, &c.; and the wet or dry state of the frustules often cause a decided difference in their appearance as regards colour. To illustrate the forms and markings of the frustules and valves, we may select the following species taking first those which occur in fresh water. In _Epithémia tur´gida_ (Pl. V. fig. 30), the side-view or valve (_s_) exhibits transverse or slightly radiating lines, with intermediate rows of dots--these markings being continued over the margins of the valves so as to appear also in the front view (fig. 30 _f_), ceasing at the hoop. The frustules are curved or arcuate (_ar´cus_, a bow) in the side view, oblong and narrowed at the ends in the front view.