Sea-sand and pebbles, washed clean, form the best bottom for the salt-water aquarium.  It must be recollected that many of the marine tenants are burrowers, and require a bottom adapted to their habits.  Some rock-work is considered essential to afford a grateful shelter and concealment to such creatures as are timid by nature, and require a spot in which to hide:  this is true of many fishes.  Branches of coral, bedded in cement, may be introduced, and form beautiful and natural objects, on which plants will climb and droop gracefully.

Sea-water dipped from the open sea, away from the mouths of rivers, is, of course, the best for the marine aquarium.  If pure, it will bear transportation and loss of time before being put into the tank.  It may, however, not always be possible to get sea-water, particularly for the aquarium remote from the seaboard, and it is therefore fortunate that artificial sea-water will answer every purpose.

The composition of natural sea-water is, in a thousand parts, approximately, as follows:  Water, 964 parts; Common Salt, 27; Chloride of Magnesium, 3.6; Chloride of Potassium, 0.7; Sulphate of Magnesia, (Epsom Salts,) 2; Sulphate of Lime, 1.4; Bromide of Magnesium, Carbonate of Lime, etc., .02 to .03 parts.  Now the Bromide of Magnesium, and Sulphate and Carbonate of Lime, occur in such small quantities, that they can be safely omitted in making artificial seawater; and besides, river and spring water always contain a considerable proportion of lime.  Therefore, according to Mr. Gosse, we may use the following formula:  In every hundred parts of the solid ingredients, Common Salt, 81 parts; Epsom Salts, 7 parts; Chloride of Magnesium, 10 parts; Chloride of Potassium, 2 parts; and of Water about 2900 parts, although this must be accurately determined by the specific gravity.  The mixture had better be allowed to stand several days before filling the tank; for thus the impurities of the chemicals will settle, and the clear liquor can be decanted off.  The specific gravity should then be tested with the hydrometer, and may safely range from 1026 to 1028,—­fresh water being 1000.  If a quart or two of real sea-water can be obtained, it is a very useful addition to the mixture.  It may now be introduced into the tank through a filter.  But no living creatures must be introduced until the artificial water has been softened and prepared by the growth of the marine plants in it for several weeks.  Thus, too, it will be oxygenated, and ready for the oxygen-using tenants.

It is a singular fact, that water which has been thus prepared, with only four ingredients, will, after being a month or more in the aquarium, acquire the other constituents which are normally present in minute quantities in the natural sea-water.  It must derive them from the action of the plants or animals, or both.  Bromine may come from sponges, or sea-wrack, perhaps.  Thus artificial water eventually rights itself.