whose direction would seem to be such that it passes up the outside of the collar, down its inner side, and out in the middle in the direction of the flagellum; and there is a distinct eddy, in which foreign particles tend to be caught, around the peripheral margin of the collar. When the cell dies, that is to say when motion ceases, the collar immediately shrivels away and disappears. It is notable, by the way, that the edge of this little mobile cup is always smooth, never notched or lobed as in the cases we have discussed on p. 236: this latter condition being the outcome of a definite instability, marking the close of a period of equiôÙlibôÙrium; while in the vibratile collar of Codosiga the equiôÙlibôÙrium, such as it is, is being constantly renewed and perpetuated like that of a jugglerãs pole, by the motions of the system. I take it that, somehow, its existence (in a state of partial equiôÙlibôÙrium) is due to the current motions, and to the traction exerted upon it through the friction of the stream which is constantly passing by. I think, in short, that it is formed very much in the same way as the cup-like ring of streaming ribbons, which we see fluttering and vibrating in the air-current of a ventilating fan.
It is likely enough, however, that a different and much better explanation may yet be found; and if we turn once more to Mr Worthingtonãs Study of Splashes, we may find a curious suggestion of