In procuring water for the aquarium, always select that which is as pure as possible. Absolutely pure water can not be had, even in the laboratory of the chemist, nor would it be desirable if it could be obtained. Distilled water is entirely unfit for starting an aquarium. The water of pure wells, lakes, and streams, is the most suitable. Some natural waters are so highly impregnated with lime, iron, or sulphur, that they are quite unsuited to our purpose, though we have seen both animals and plants thriving in such waters. It will be found, however, on examination, that these plants and animals have become acclimated, - as it were, reconciled to their conditions; and also that there are at work certain countervailing influences which we may find it difficult to imitate. Whenever we have attempted to use such water, - and we have frequently done so for experiment, - all plants and animals not born and brought up in it have suffered.
Water impregnated with iron or sulphur is, in general, quite local in its occurrence. A certain pond or stream will be strongly tainted, and streams only a few yards distant will be quite pure. But in many districts of country, all the water contains such a large percentage of lime that it is unfit for aquarium purposes. In such cases recourse should be had to rain-water, caught at a distance from houses and well filtered. With such water we have succeeded admirably.
After the aquarium has been filled and the plants have begun to grow nicely, it will be found that the water gradually diminishes on account of evaporation; and this is specially marked in those aquaria that are kept in warm rooms, where the air is dry. In such cases it will be found that it is the water alone that evaporates and is wasted; the salts and other impurities remain behind. If we now fill up the tank with water, such as we originally used, and which contains the same amount of saline matter that the water did, it is evident that we add to the original impurities; and by keeping up this practice, we will soon have more salts present than is endurable. This is particularly the case with marine aquaria: if we keep on adding sea-water to make up for the evaporation, it will soon attain a Dead-Sea degree of saltness. To avoid this we must simply imitate Nature, and make up for the loss by evaporation by adding rain-water, - which should, however, for aquarium purposes, always be well filtered. At proper intervals - three months or so - we should draw off a large proportion of the water in the tank, and fill up with newly collected water, - fresh or salt, as the ease may require.
It will sometimes be found that certain plants and animals whose habitat is boggy and impure water will not thrive well in any other. In such cases, the only way to secure success is to set up a separate tank, in which the natural conditions are imitated as closely as possible.