tion will suffice for the production of a new individual; a single tentacle will produce a flower and stem, and finally a whole colony. A transverse section of the stem will produce a flower at the distal end, and a continuation of the stem, with the process by which it attaches itself, at the proximal end of the section. Just so far it shows orientation—that the stem has a distal and proximal end. There is no sign of bilaterality in most species, and in others the indication is so slight that it is hardly worthy of the name. This development of the flower always at the distal, and of the stem always at the proximal, extremity of the section, shows conclusively that the stem grows both ways, and that in every segment there exists a neutral plane midway between both ends.
Besides these plant-like modes of reproduction, hydroids are generated, like the actiniæ, by spontaneous fission, a development of one individual into two or more by a natural vertical cleavage.
They multiply by ova, by ovules, by independent ciliated embryos, like the lower invertebrates, the reptiles, and birds. Some varieties possess a sort of marsupial pouch, in which the undeveloped young
are retained till they attain maturity; and, like the mammals, in some cases, the individual quits the parent after attaining perfect development. Added to all these modes of reproduction, in which the analogy must not be pressed too closely to those of higher organisms, they possess two very curious modes of their own; one given by Allman in his monograph, the other by Carpenter in the latest edition of "The Microscope, and its Revelations." The Tubularian and Campanularian hydroids, Allman tells us, develop upon their stems bell-shaped medusæ (Figs. 4, 5, 11), which free themselves and swim the adjacent waters. All free-swimming medusæ have not yet been traced to hydroid stems; but, as all which have been carefully studied through their life-history are found to originate there, it is supposed to be true of the others.