Assuming the process in Pyrosoma to be relatively primitive, it would seem that budding has much in common with regenera tion. In Amphioxus the various tissues have a certain power of self-regeneration, like repairing like, but incapable of more (Andrews, Franz). Many small Ascidians undergo periodically a process of thoracic degeneration followed by repair, which implies the retention of undifferentiated material at the base of pharynx and atrium, where growth normally takes place most actively (cf. Diazono-Aphanibranchion, Oka). We may safely assume that in primitive Tunicates thoracic degeneration was a normal phenomenon, and that the machinery for rapid repair took the common form of regenerative disks or pockets in these places. But there is a certain antithesis between regeneration and bud ding. Only when budding is slight is regeneration useful: the repair of the individual ceases to be important when new indi viduals are turned out rapidly by mass-production.
It thus seems possible to account for the variety in the modes of Tunicate budding by assuming that the primitive Tunicate possessed a simple stolon like that of Cephalodiscus and regener ative pockets at the base of pharynx and atrium. As the special isation of the ectoderm for test-production tended to disqualify the ectoderm from playing its original major part in the budding process (e.g., Cephalodiscus), the descent into the stolon of ex tensions from the regenerative pockets would produce a complex stolon like that of Thaliacea, with the result of speeding up the rate of budding and reducing the value of mere regeneration. The elimination of one or other pair of stolonial elements in Ascidiacea may be readily associated with the changes already described in their respective attitudes after metamorphosis.
The organisation of the buds to form locomotive colonies in the Thaliacea has been outlined in the systematic part. Both in Pyrosoma and the wheel-colonies of Cyclosalpa the integration reaches a notable degree, even though in the latter case the zooids remain capable of a free life after break-up of the colonies. The following early account of the wheels of C. pinnata is worth saving from oblivion: "A beautiful sight, these rings of a dozen gelatinous finger-long individuals, all set vertically with their mouths above (and with their ventral 'ins' converging to the centre like the spokes of a wheel), swimming about with rhythmic motions as if controlled by a common will, now gently rocking, now twisting and turning, or steering straight on through the quiet water" (translated from Carl Vogt, Bilder aus dem Thier-leben, 8 5 2 ) .
Equally noteworthy are the arrangements in Doliolids by which clusters of small amoeboid "phorocytes" take possession of the small buds, as they are detached from the ventral stolon, and transport them to their places on the dorsal outgrowth. In Pyrosoma, where the conditions are simpler, these phorocytes resemble the ordinary stellate and spindle-shaped "test-cells," and must be regarded as test-cells devoted to this particular task. In Doliolum they are proliferated in great numbers by the ecto derm at the base of the stolon (Neumann), although the test itself is homogeneous and ordinary test-cells are absent. The remarkable capacity of some Tunicates to return to a less differ entiated condition is discussed in detail in the article DEDIFFEREN TIATION.