This section is from the book "The Goldfish And Its Systematic Culture With A View To Profit", by Hugo Mulertt. Also available from Amazon: The goldfish and its systematic culture.
As the goldfish deposits its spawn upon plants that live in the water, it is necessary that the natural condition of things be closely imitated in preparing the bed in which the fish are expected to spawn. These plants can be obtained from any neighboring creek or marsh (that with a gravel bottom preferred). Those marsh plants possessing fibrous roots are either pulled or dug up with their roots entire; they are then washed thoroughly to cleanse them of the adhering mud, and closely examined to see that there are no eggs of other fish or insect larvae upon them. Having secured enough of these for the present, say half a dozen good clumps for each bed, they are then placed loosely in the water of the bed, along that side where the sun shines upon them in the morning. On these roots the females will deposit their eggs Later in the season, when aquatic plants have commenced to grow, these roots can be removed and replaced with such aquatics as the Horn or Waxworth (Ceratophyllum demersum), and the Canal Pest (Anacharis canadensis). These plants are especially good to catch the eggs when dropped by the fish.
The water supply is now shut off from the bed, only an occasional supply being let in to preserve the proper height at which the water should remain.
If the goldfish are kept in places where it is impossible to get them out, and their spawn is wanted, large bunches of the fibrous roots can be tied to a string fastened to the shore, and allowed to float about upon the water. When containing spawn, they can be brought out merely by pulling the string, when the eggs can be removed and taken care of by the one in charge.
When the fish are wintered over in sheltered localities, as a greenhouse for instance, the spawning season will commence a great deal in advance of that outside. In such a location, spawn may be expected from the middle of February, while that event seldom takes place in the open air earlier than the latter part of April or the beginning of May. When the spawning has begun, it is continued until the setting in of frost, with now and then an interval of a week or two.
In northern climes, where the season is short, it may be lengthened by placing over the spawning bed and nursery frames covered with glass, the principle being the same as that of the gardener's hot-bed. If this is done, it must not be forgotten that in fair weather an abundance of air should be admitted, and also that the bed is to be protected from great and sudden changes of temperature, as either would be fatal; in fact, the breeder is to exercise his intelligence in the matter, for he certainly must know that as he is keeping up a kind of artificial climate, he must not forget to preserve it as evenly as possible, otherwise one cold night would (if the frame had not been replaced after airing the bed) destroy many days of patient toil.
If, in between spawning periods, the fish should rest longer than is desired by the breeder, the males are to be changed from one bed into the other, and the water in the latter well aerated by letting a stream flow through it for a whole day, when the spawning will soon be in full process again.