[Illustration: FIG. 71.—Portion of the skeletal framework of _E. fluviatilis_. _a_, Main fibres; _b_, connecting fibres. (After Weltner.)] [Illustration: FIG. 72.—Spicules of _E. fluviatilis_. _a._ _b._ _c._ Oxeas, spined and smooth; _d._ _e_, amphidiscs, side and end views. (After Potts.)] Weltner[205] has shown that on the death and disintegration of the mother sponge some of the gemmules remain attached to the old skeleton, some sink and some float. Those which remain attached are well known to reclothe the dead fibres with living tissue. They inherit, as it were, the advantages of position, which contributed to the survival of the parent, as one of the selected fittest. The gemmules which sink are doubtless rolled short distances along the bottom, while those which float have the opportunity of widely distributing the species with the risk of being washed out to sea. But even these floating gemmules are exposed to far less dangers than the delicate free-swimming larvae, for their soft parts are protected from shocks by a thick coat armed with amphidiscs. The gemmules are likewise remarkable for their powers of {178}resistance to climatic conditions, powers which must contribute in no small way to the survival of a species exposed to the variable temperatures of fresh water. Thus, if the floating gemmules or the parent skeleton with its attached and dormant offspring should chance to be included in the surface layer of ice during the winter, so far from suffering any evil consequences they appear to benefit by these conditions. Both Potts and Weltner have confirmed the truth of this statement by experiments. Weltner succeeded in rearing young from gemmules which had suffered a total exposure of 17 days to a temperature "under 0° C." Of important bearing on the question of the utility of the gemmules are certain instances in which _E. fluviatilis_ has been recorded as existing in a perennial condition.[206] The perennial individuals may or may not bear gemmules, which makes it evident that, with the acquisition of the power to survive the winter cold, the prime necessity of forming these bodies vanishes. The perennial specimens are described as exhibiting a diminished vegetative activity in winter, the flagellated chambers may be absent (Lieberkühn), or present in unusually small numbers (Weltner), the entire canal system may be absent (Metschnikoff), or, on the other hand, it may be complete except for the osculum. [Illustration: FIG. 73.—Gemmule of _E. fluviatilis_. _b_, Amphidisc. (After Potts.)]