Films ofPimpla eggs in which three different density gradients were produced (Nuss, 1974) made it possible to analyse the dynamic conditions that form the basis for the correct course of cleavage, of ooplasm movements and the distribution of energids. The results of sublethal stresses on the eggs led to the postulate that there is a system of factors which are arranged in a cylindrical wall in the central plasm flowing system, linked with the marginal streaming in opposite direction (Bruhns, 1974). Centrifuged eggs developing normally do not show any visible change in longitudinal streaming or in energid migration, as a result of the different stratification after spinning or after swinging at a short distance as well further from the rotational axis. Ooplasmic movements can take place independently of the arrangement of shiftable constituents visible in the light microscope. From the means of cleavage aberrations that result from centrifugation three components of the supporting and moving system can be distinguished. 1. The submicroscopic longitudinal flowing mechanism can not be shifted in a lateral direction within the egg by density gradients. It corresponds to the potential cylindrical system of factors which is either hindered or able to be restired after accelerations above 5800 g but which is destroyed when the whole egg is stressed at 10000 g. In this case, the eggs do not show any longitudinal flowings and their ingredients shift according to the gravitational forces. The flowing mechanism helps to, also supporting the normal egg architecture. After the egg has been activated, during the mixing motion and the formation of periplasm the flowing mechanism has to be completed in structure and put in working order for the unipolar flows. In centrifuged eggs, the initial region of the fountain flow may be shifted along the egg axis or it may not be distinctly visible, nevertheless normal larvae appear. 2. When the whole egg architecture has been overstressed, polar caps do not appear and those already formed during the unipolar flows are lost. This contraction of ooplasm may happen not, missing independently of the flowing mechanism, so that we have to postulate a specialmechanism of ooplasmic contraction. It should be found near the egg surface. The annular ooplasm contractions which appear later in cleavage might be in synergistic connexion with the bipolar fountain flows. 3. The flowing mechanism and the mechanism of contraction can function independently of the presence of energids. On the other hand the energids are able to migrate independently of flows. An essentialmechanism of energid migration might be the "migration aster" (Wolf, 1973); by its means the nucleus is transported and its cytoplasmic island can be extended. The redistribution begins just where energids migrate, multiply and from new cytoplasmic islands, i.e. where ooplasm is converted. In a caudal half of the egg without fountain flow the nergids are able to go in all directions and to reach the egg surface directly way. 4. The distinct formation of an anterior (Fz), a medium (Dz) and a posterior initial region (Bz) of ooplasm flows in centrifuged eggs is discussed. The unipolar flows appear to be necessary to lead the pronuclei into the cleavage center (Fz). Energid migration which is activated, is limited to a small region of converted ooplasm which is transported as a whole by transfer flow into differentiation center (Dz). If energids remain along this route the cephalic fountain does not appear: meanwhile the caudal one transports the energids from theDz towards the caudal end of the egg. The flowing mechanism must be a barrier to migration asters so that they are not able to extend in a radial direction. But also towards the poles they are coordinated to form a streaming front, for in eggs with damaged energid systems, corresponding flow resistance regions are formed. In consequence of the longitudinal flowing mechanism the cleavage energids, when migrating in a radial direction, react at about the same time as the marginal plasm and form nucleoli. Because normal larvae hatch from eggs with an aberrant course of cleavage, energids in the central plasm must be isopotent.