The energy of movement is generated in the moving layers themselves, and these are retarded by friction against the non-moving ectoplasm to an extent determined by their own viscosity, and to a much less extent by friction against the cell-sap which is passively carried with the stream. The motor-mechanism is such that no backward reaction is exercised upon the external or internal layers. The velocity of streaming is largely dependent upon the viscosity of the protoplasm, and hence also upon the percentage of water in the latter, but the osmotic pressure exercises little or no direct influence upon streaming.
The activity of diosmosis is not necessarily increased by the existence of streaming, but secondarily induced differences of osmotic pressure may be perceptible between streaming and quiescent cells.
Albuminous solutions containing 89 90 per cent, of water have at 18 20 C. a viscosity of from O'OG to O07, with 95 per cent, of water 0-04, with 72 per cent. 0-29 C.G S. unit.
Gravity exercises little or no influence upon streaming in small cells, and only a very slight indirect action on streaming in large ones. The velocity of floating particles of greater or less density than the plasma may be distinctly affected by gravity, a fact which indicates that the viscosity of the streaming plasma is comparatively low.
As the temperature rises within certain limits (0 C. to 45 or 50 C.), the viscosity of the plasma decreases, and a large part of the increased velocity is due to this cause alone.
Assuming that the viscosity of the endoplasm corresponds approxi- mately to that of 10 per cent, albumin, forces of 8*75 and 219 dynes would be required to impart velocities of 2 mm. and 0'4 mm. per minute to a gramme of moving liquid in cells of 0*1 and O'Ol cm. internal diameter. The amount of work done in a year represents, in the first case, a theoretical consumption of only 77^075^11 of a gramme of cane-sugar per gramme of moving liquid, and in the second repre- sents only TuVtfth of the energy of respiration, even if 99 per cent, of the energy directed towards streaming is wasted. The force required varies inversely as the square of the radius, so that in a 50-cm. tube of the cribral system of Cucurlrita, having 2000 sieve-plates with sieve- pores of 2 /A diameter and 10 /* length, a pressure of approximately half an atmosphere would be required to produce a movement in mass of the contents through the tube at a rate of 0-5 mm. per minute. Hence an internal pressure of 5 atmospheres would cause an escape of the contents of an opened sieve-tube at an average rate of about 5 mm. per minute, so long as the pores remained unblocked, which corresponds with the slow exudation taking place when a mode- rately turgid sieve-tube is first opened.
Regular protoplasmic streaming does not seem to occur even in the individual segments of sieve-tubes, and hence there is no need to discuss whether the protoplasm could generate the propulsive force
