The marine fifteen-spined stickleback (_Gasterosteus spinachia_) affords another instance of nest-constructing fishes. The places selected for their nests are usually harbours, or some sheltered spots to where pure sea water reaches. The fish either find growing, or even collect some of the softer kinds of green or red seaweed, and join them with so much of the coralline tufts (_Janiæ_) growing on the rock as will serve the purpose of affording firmness to the structure, and constitute a pear-shaped mass five or six inches long, and about as stout as a man's fist. A thread, which is elastic and resembles silk, is employed for the purpose of binding the materials together: under a magnifier it appears to consist of several strands connected by a gluey substance, which hardens by exposure to the water.[113] M. Carbonnier, who has studied the habits of the Chinese butterfly-fish (_Macropodus_) in his private aquarium in Paris, where he had some in confinement, observed that the male constructs a nest of froth of considerable size, 15 to 18 centimetres horizontal diameter, and 10 to 12 high. He prepares the bubbles in the air (which he sucks in and then expels), strengthening them with mucous matter from his mouth, and brings them into the nest. Sometimes the buccal secretion will fail him, whereupon he goes to the bottom in search of confervæ, which he sucks and bites for a little in order to stimulate the act of secretion. The nest prepared, the female is induced to enter. Not less curious is the way in which the male brings the eggs from the bottom into the nest. He appears unable to carry them up in his mouth; instead of this, he first swallows an abundant supply of air, then descending, he places himself beneath the eggs, and suddenly, by a violent contraction of the muscles in the interior of his mouth and pharynx, he exhales the air which he had accumulated by the gills. This air, finely divided by the lamellæ and fringes of the gills, escapes in the form of two jets of veritable gaseous powder, which envelopes the eggs and raises them to the surface. In this manoeuvre the _Macropodus_ entirely disappeared in a kind of air-mist, and when this had dissipated he reappeared with a multitude of air-bubbles like little pearls clinging all over his body.[1] Again, in detailing Mr. Baker's observations on the three-spined stickleback, published in the Philosophical Transactions, this author says:--