All the many hundreds of species of Polychaetes are marine, with a very few exceptions, which have been in recent years recorded from fresh (_i.e._ drinkable) water, viz. a species of _Nereis_ from a lake in Mingrelia, another _Nereis_ and a _Lumbriconereis_ from running water in Trinidad,[340] a Sabellid, _Manayunkia speciosa_,[341] from Philadelphia; and another Sabellid, _Coabangia_,[342] from fresh water at Tonquin, which lives in borings in shells of _Melania_; and it is by no means improbable that other fresh-water Polychaetes exist in Lake Tanganyika in Africa, where a Medusa has recently been discovered. In brackish water of various densities many Polychaetes live; _Arenicola_ especially is regardless of the character of the medium, and _Nereis diversicolor_ appears to withstand considerable admixture of fresh water. The majority of the Polychaetes occur "inshore," that is, between tide-marks and in shallow water down to 20 fathoms; but they occur at all depths more or less abundantly, and some have been dredged from depths of more than 3000 fathoms. The nature of the soil composing the shore has a good deal to do with the number of worms to be found there; thus in calcareous districts they are fewer than in places where harder rocks, {285}such as granite, form the shore line, for the chalk or limestone wears away more quickly, and exposes to destruction the worms which may have sheltered in its crevices: further, it does not give so permanent a place of attachment to seaweeds, on which many Polychaetes feed. The calcareous rocks, too, are more likely to be traversed by springs of fresh water, which is not to the taste of the worms. The sand resulting from the destruction of the rocks, whether hard or soft, is of itself unsuitable to the majority of worms, which are most abundant where mud containing decaying vegetable matter is mixed with the sand: this, which gives a firmer consistency to the soil, so that the burrows retain their form better, supplies food for the burrowers. GENERAL HABITS.—The division of the Polychaetes into the "Errantia" or free-swimming and wandering forms, and "Sedentaria" or tubicolous and sedentary forms, is a misleading mode of classification, for as a matter of fact only a comparatively few forms are really free-swimming throughout life; the majority, even if they do not form definite tubes, burrow galleries for themselves in the soil, and these burrows are in many cases only rarely left; this is true of both groups. Amongst the "errant" Polychaetes nearly all the Eunicidae secrete a parchment-like tube, and some Polynoids form mud tubes. Among the "sedentary worms" there are forms which merely burrow; while _Myxicola_ readily leaves its gelatinous tube and swims freely; _Pectinaria_ carries its house with it as it moves about, and _Polycirrus_, a Terebellid, does not form any tube at all.