Disappointed, but not discouraged, I filled several shallow pans with pure sea-water, clean sand, and fresh plants, and transferred to them my suffering and wellnigh exhausted animals.  A day restored them to their normal condition, and now I was ready to begin my Aquarium anew.

But to what purpose should I begin anew?  Would there not be the same failure?  What had been wrong?

At least two great faults were evident.  First, in order to guard against the possibility of a leak, the bottom and posts of the tank had been covered with many coats of an alcoholic varnish.  Now it was probable that time enough had not elapsed between the several applications for the thorough evaporation of the alcohol.  Might not its gradual infusion in the water have caused the death of the animalcula in such numbers as to taint the whole by their decay?

The second fault was, strewing upon the surface of the sand a handful or two of white powdered quartz, which, from having been pulverized in an iron mortar, was so oxydized as to turn a deep yellow.  This might have poisoned the animalcula.

The first fault seemed to me the chief, but I proceeded to remedy both.  The whole contents of the tank being removed, it was thoroughly washed on the inside, exposed for several days to the sun and air, and then soaked for twelve hours in clean sea-water.  This being thrown away, the stones, scalded and well-washed, were restored, and clean sand, replaced the old.

Water was drawn from the dock at high tide; but it was less clear now, on the fourth of June, than that which had been got early in May.  This surprised me not a little; for, as I stood upon the wharf and looked down into it just before sunset on the previous evening, I was struck with its beautiful limpidity.  Curious to see if its aspect remained unaltered, I went to the same spot where I had stood the night before.  The tide was at the same height, but twelve hours had made a marvellous change in the appearance of the water.  Its sparkling clearness had given way to greenness and turbidity, and no object could be seen a foot below the surface.  No storm had stirred its depths during the night,—­why this change?  Conjecture was of no practical utility, and I returned home satisfied that my fifteen gallons of water were as clear as any it was then in my power to obtain.  Covering the tub from the dust, I left it to settle until sunset.  Then the ever-useful siphon drew off two thirds of it tolerably clear, leaving a thick green deposit upon the sides and bottom of the vessel.  Next day, it was again drawn off from the sediment, (at this time, small in quantity,) and poured into the tank.  Several newly obtained plants of well-growing Enteromorpha and Corallina were arranged among the stones, and the Aquarium was left at rest.  Gradually the water became nearly clear, but not perfectly so until after the introduction of animals.

Eight days after it was filled, the Actinias were put in; on the ninth, several small Mollusks; on the tenth, Crustacea; and on the eleventh and twelfth, other varieties of the same types; but not until the fourteenth day were fishes ventured upon.