Concerning the first of these we must caution the reader against the common error of overcrowding the aquarium with animals. It must be remembered that almost all marine animals obtain the oxygen gas required for purposes of respiration from the air dissolved in the water. Now, atmospheric air is only very slightly soluble in water, and hence we can never have an abundant supply in the water of an aquarium at any one time. If a number of animals be placed in any ordinary indoor aquarium, they very soon use up the dissolved oxygen; and, if no means have been taken to replace the loss, the animals die, and their dead bodies soon begin to putrefy and saturate the water with the poisonous products of decomposition. It is probably well known to the reader that a large proportion of the oxygen absorbed by the respiratory organs of animals is converted by combination of carbon into carbonic acid gas within their bodies, and that this gas is given back into the water where it dissolves, thus taking the place of the oxygen used in its formation. If, then, an aquarium of any kind is to be a success, some means must be taken to keep the water constantly supplied with fresh oxygen quite as rapidly as it is consumed, and this can be done satisfactorily by the introduction of a proportionate quantity of suitable living weeds, providing there is not too much animal life present. The majority of living plants require carbonic acid gas as a food, and, under the influence of light, decompose this gas, liberating the oxygen it contained. This is true of many of our common sea-weeds, and thus it is possible to establish in a salt-water aquarium such a balance of animal and vegetable life that the water is maintained in its normal condition, the carbonic acid gas being absorbed by the plants as fast as it is excreted by animals, and oxygen supplied by the plants as rapidly as it is consumed by the animals. This condition, however, is more difficult to obtain in a salt-water aquarium than in one containing fresh-water life, partly because, generally speaking, the sea-weeds do not supply oxygen to the water as rapidly as do the plants of our ponds and streams, and partly because of the difficulties attending the successful growth of sea-weeds in artificial aquaria. Thus it is usually necessary to adopt some means of mechanically aërating the water; but, for the present, we shall consider the sea-weeds only, leaving the mechanical methods of aërating the water for a later portion of this chapter.