required to produce moderately rapid streaming through the pores. The differences of hydrostatic pressure between the different seg- ments of an intact sieve-tube would, however, frequently suffice to produce a direct transference in mass of the fluid contents through the pores of sieve-tubes at a millimetre or so per minute,* for, as evidenced by the influence of gravity and of centrifugal force upon suspended particles, the viscosity of the contents of the sieve-tubes is relatively low.
In the case of bacterium cells of 2 to - 5 /A diameter, it is very doubtful whether the protoplasm could with economy generate the required propulsive force for perceptible internal streaming movements. The latter take place in diatoms mainly on the external surface, and require relatively little expenditure of energy.
In the case of protoplasmic threads passing through fixed channels in the cell-wall, if these are y^ //. diameter and 5 /A length, a pressure of over 6 atmospheres would be required to drive a liquid of viscosity O075 through a single thread at a velocity of 1 mm. per second. So that if a tissue-cell were isolated, it would at the lowest computation take several years for the escape of 1 cubic mm. of the cell-contents through one such thread under a maximal internal osmotic pressure, even sup- posing that the thread did not become plugged by floating particles and that no cellulose was formed across its exposed end. The outer layers of the ectoplasm appear, moreover, to be very much more viscous than the endoplasm, and hence it is obvious that the threads do not subserve purposes of translocation, although minute particles of protoplasm may be bodily transferred from one cell to another in the course of time.
The direction of streaming is mainly determined by internal factors, and in rotating cells a reversal is only possible in certain cases and under very special conditions.
Changes occur spontaneously however in cells exhibiting circulation. The total resistance during circulation is greater than during rotation, and hence, unless the velocity increases considerably, a change from the former to the latter after stimulation is not due to an increased energy of streaming but to a change in the configuration of the protoplasm.
The energy for streaming can be derived either from aerobic or anaerobic metabolism. Certain species of Cham and Nitella are in fact facultative anaerobes, and may exhibit slow streaming for 6 8 weeks in the entire absence of free oxygen.
No special chemical changes are connected with streaming.
- .According to de Bary (' Compt. Anafc.,' 1884, p. 177) finger-like processes
from the adjoining segments of a sieve-tube meet in the sieve-pores, but remain distinct. This discontinuity lias only been observed in dead sieve-tubes, and it probably results from the breaking of the viscous protoplasmic threads at their thinnest points on death, which is a surface-tension effect commonly produced in dying protoplasmic threads.
