The actual mode of formation of the epiphragm would seem to differ in different species. According to Fischer,[43] the mollusc withdraws into its shell, completely blocking all passage of air into the interior, and closing the pulmonary orifice. Then, from the middle part of the foot, which is held exactly at the same plane as the aperture, is slowly secreted a transparent pellicle, which gradually thickens, and in certain species becomes calcareous. Dr. Binney, who kept a large number of _Helix hortensis_ in confinement, had frequently an opportunity of noticing the manner in which the epiphragm was formed.[44] The aperture of the shell being upward, and the collar of the animal having been brought to a level with it, a quantity of gelatinous matter is thrown out [? where from]. The pulmonary orifice is then opened, and a portion of the air within suddenly ejected, with such force as to separate the viscid matter from the collar, and to project it, like a bubble of air, from the aperture. The animal then quickly withdraws farther into the shell, and the pressure of the external air forces back the vesicle to a level with the aperture, when it hardens and forms the epiphragm. In some of the European species in which the gelatinous secretion contains more carbonate of lime, solidification seems to take place at the moment when the air is expelled, and the epiphragm in these is in consequence strongly convex. =Thread-spinning.=--A considerable number of fresh-water Mollusca possess the power of stretching a thread, which is no more than an exceedingly elongated piece of mucus, to the surface of the water, and of using it as a means of locomotion. This thread bears no analogy whatever to the fibrous byssus of certain bivalves, being formed in an entirely different manner, without the need of a special gland. The threads are ‘spun’ by several species of _Limnaea_, _Physa_, and _Planorbis_, by _Bithynia tentaculata_, and several of the _Cycladidae_. They are anchored to the surface by a minute concavity at the upper end, which appears to act like a small boat in keeping the thread steady. The longest threads are those of the _Physae_, which have been noticed to attain a length, in confinement, of 14 inches. They are always spun in the ascent, and as a rule, when the animal descends, it rolls the thread up and carries it down as it goes. A single thread is never spun on the descent, but occasionally, when a thread has become more or less of a permanence, it becomes stronger by the addition of more mucus each time it is used, whether for ascending or descending purposes. _Cyclas cornea_ appears to be an exception to the rule that threads are only spun on the ascent. This species, which is particularly fond of crawling along the under surface of the water, has been noticed to spin a thread half an inch in length while on the surface, and to hang suspended from it for a considerable time.