The development of a new part at right angles to an oblique surface has also been described, and it has been pointed out that the result appears to be due to the symmetrical development of the new structure in the new part. This symmetry of the newly forming part must be also counted as one of the properties of the organization. Finally, the mode of regeneration of a new, bifurcated tail in the teleost, stenopus, shows that the new part may very early become moulded into the characteristic form, and that the growth of the different parts is regulated by the structure assumed at an early stage. The new part does not grow out at an equal rate until it reaches the level of the notch of the old tail, and then continue to grow at two points to produce the bilobed form of the tail; but the bilobed condition appears at the very beginning of the development. These illustrations give us nearly all the data that we possess at present on which to build up a conception of the organization. That we must fail in large part fully to grasp its meaning from these meagre facts is self-evident. The main difficulty seems to lie in this,--that when we attempt to think out what the organization is we almost unavoidably think of it as a structure having the properties of a machine, and working in the way in which we are accustomed to think of machines as working. The most careful analysis of the “machine theory,” as applied to the phenomena of development and of regeneration, has been made by Driesch. It has been pointed out that in his _Analytische Theorie_ Driesch assumed that development is due to “given” properties in the egg; that each stage is initiated by some substance contained in the egg acting on the stage that has just been completed. That is, each stage is the condition of the following. The “rhythm” of development is accounted for in this way. The changes are described as due to chemical processes (including also ferment actions). The nucleus is supposed to contain all the different kinds of ferments that act, when set free, as stimuli on the protoplasm; but since the ferments are always set free at the propitious moment, Driesch was obliged to assume that the cytoplasm acts on the nucleus in such a way as to make it produce the proper ferment for the next stage. Thus the cytoplasm first influences the nucleus, the latter sets free a specific ferment that starts a new chemical change in the cytoplasm, and the changed cytoplasm may then react again on the nucleus, and a different ferment be set free, etc. Each change is therefore not only an effect of what has gone before, but the cause of the next process.[132] Driesch points out that it is necessary at this stage to make a further assumption, because the cytoplasm must not only be acted upon by the ferment, but it must itself be of such a sort that it _responds_ to the action. This leads to a great complication of the phenomena; but the assumption does not depart,