[Illustration: Figure 32. Showing ingestion of alga filaments in _Frontonia leucas_. _a_, the beginning of the ingestion of an alga filament. Note the streaming of the endoplasm _preceding_ the end of the filament. _b_, almost two complete coils of the filament have been rolled up inside the _Frontonia_ by the rotary streaming endoplasm. The endoplasm in the center of the animal is stationary. _c_, a filament, if thin, may be grasped anywhere along its length, bent together and swallowed in the usual manner. Diameter of _a_, 250 microns.] If a filament of alga is too long for the _Frontonia_, or one end of it is fast, streaming is reversed after several coils have been rolled up and the filament is ejected. So far as could be observed, the streaming process is reversed in all details, though the rate of ejection seemed to be somewhat slower than the rate of ingestion. Occasionally, however, ejection is accomplished much more quickly. If there are several coils of a filament whose other end is fast, rolled up inside of a _Frontonia_, the mouth sometimes stretches antero-posteriorly until the coil as a whole without unwinding is thrown out of the body. The viscosity of the endoplasm might lead one to expect that some of the endoplasm would be brought out with the alga, but such is not the case. The essential differences between rotational streaming in _Frontonia_ and in paramecium are: (1) It is under the control of the organism in _Frontonia_ while in paramecium it is a continuous reversible process. (2) It is much more rapid in _Frontonia_ than in paramecium. On the other hand, the physics of streaming in both organisms is essentially the same, so far as could be detected. In both organisms the energy of streaming is liberated within the endoplasm. This is especially well shown in the first stages of feeding. Besides these organisms in which streaming occurs, either in a part of the organism or the whole, streaming is also found to occur in a great variety of plants other than those already mentioned; in the leukocytes of perhaps all coelomates; in some animal egg cells, such as the sponges, hydra and molluscs; in pigment cells, especially in batrachians and lacertilians; in phagocytes and wandering cells of a great many animals; in the nuclei of some animal cells; and in the intestinal epithelial cells of perhaps all metazoans. In almost none of these cases however do we know more than the bare fact that streaming occurs. No details are known. Consequently in so far as the purposes of this book are concerned it will not be apropos to discuss these cases further except to record the thesis that there is no evidence tending to show that these cases are not at bottom all characterized by the operation of the same fundamental process.