loosing itself in its upper part into the capillary point, which is often very long, and has very thin articulations. When a filament has left by creeping the mother’s body (as in fig. 72), this content (b.) forms a point, often similar to a drop, which grows gradually longer, till it acquires the form of an articulated filament; and in this manner the process already described is successively performed.
Other Oscillatoriae (such as the Oscillatoria elegans Ag. fig. 73—74) creep in turning and bending the pointed extremity of their head (a.), from upwards downwards, or from right to left, as the movement requires.
If, during this movement, filaments or other exiguous bodies present themselves on its way, the animalcule feels them with the point of the head, till it has gone round or over them. Does it meet larger bodies, it adheres to them or ascends over them, but it very soon puts aside the smaller ones, or, as the Naviculae do, drives them along with itself.
The Frustuliae move in less than a minute, upon a surface of one line. The Oscillatoriae, in four and twenty hours, move from one to four inches.
Our Diatomeae, found in rivers, are almost motionless; those which inhabit the sea move quicker. Many animalcules of this class move only during the night. I have seen the Closterium costatum (fig. 61.) and the Closterium didymotocum (fig. 64.) ascend, during that time, from the bottom of a glass cylinder, from 4—30 lines, upon the inside of the glass, whilst,
