Speculations as to the ancestral connection of the great group of vertebrates with other great groups have been varied and ingenious; but most naturalists are now inclined to the view that it is a mistake to assume any such connection in the case of vertebrates of a more definite character than we admit in the case of starfishes, shell-fish, and insects. All these groups are ultimately connected by very simple, remote, and not by proximate ancestors, with one another and with the ancestors of vertebrates. [Illustration: FIG. 21. The unicellular parasite _Benedenia_, from the gut of the common Poulp or Octopus. 1 is the normal male individual; 2 and 3 show stages in the production of spermatozoa on its surface by budding; 4, 5 and 6 show a female parasite with spermatozoa approaching it.] [Illustration: FIG. 22. Production of spermatozoa on the surface of the unicellular parasite _Coccidium oviforme_, from the Rabbit’s intestines.] The origin of the limbs of vertebrates is now generally agreed to be correctly indicated in the Thatcher-Mivart-Balfour theory to the effect that they are derived from a pair of continuous lateral fins, in fish-like ancestors, similar in every way to the continuous median dorsal fin of fishes. [Illustration: FIG. 23. Spermatozoa (often called “microgametes”) of the unicellular parasite _Echinospora_ found in the gut of the small Centipede _Lithobius mutabilis_.] The discovery of the formation of true spermatozoa by simple unicellular animals of the group Protozoa is a startling thing, for it had always been supposed that these peculiar reproductive elements were only formed by multicellular organisms (figs. 21, 22, and 23). They have been discovered in some of the gregarina-like animalcules, the _Coccidia_, and also in the blood-parasites.