From what has been said the reader will see that the total quantity of water required in this instance need not exceed the capacity of the aquarium; also that each of the vessels connected with water supply and waste should have a capacity equivalent to the volume of water contained in the aquarium between the two levels _h_ and _l_. The alternate rising and falling of the water produced in the manner just described represents in miniature the flow and ebb of the tides, but perhaps this is in itself of no great advantage in the aquarium except from the fact that it allows those creatures that prefer to be occasionally out of the water for a time a better opportunity of indulging in such a habit. And further, with regard to both the arrangements for aëration above described, it should be noted that earthenware vessels are much to be preferred to those made of metal for the holding of sea water, since the dissolved salts corrode metallic substances rather rapidly, and often produce, by their chemical action, soluble products that render the water more or less poisonous. [Illustration: FIG. 39.--AQUARIUM FITTED WITH APPARATUS FOR PERIODIC OUTFLOW] Other methods of aërating the water of aquaria are practised, but these, as a rule, are only practicable in the case of the large tanks of public aquaria and biological laboratories, as the mechanical appliances necessary to carry them out successfully are beyond the means of an ordinary amateur. In such large tanks as those referred to it is common to force a fine jet of air into the water by machinery. Sometimes this air is driven downward from a jet just below the surface, and with such force that a multitude of minute bubbles penetrate to a considerable depth before they commence to rise, but in others the air is made to enter at the bottom and must therefore pass right through the water. Of course the amateur aquarium-keeper may carry out this method of aëration with every hope of success providing he has some self-acting apparatus for the purpose, or can depend on being able himself to attend to a non-automatic arrangement at fairly regular intervals, always remembering that a single day’s neglect, especially in the case of a small tank with a proportionately large amount of animal life, may lead to a loss of valuable specimens.