MOVEMENTS OF GRANULES 319 



membrane, bordering on the protoplasm, while in its external 

 zone, as is equally necessary, a stream runs forward to the 

 anterior end, which is alone that which makes itself felt in the 

 surrounding water, and thus causes the current that is so 

 strangely reversed in the latter. It is a question whether we 

 are justified in assuming double currents, running one upon the 

 other, in a membrane of such extreme tenuity as that postulated 

 must be. In this respect I would like just to refer to the 

 following observation, which I made as far back as the seventies 

 on the occasion of studies upon cell division, when I was induced 

 to occupy myself a good deal with phenomena of surface tension. 

 In the extremely thin membrane of large soap-bubbles it is 

 possible to produce very violent streaming movements by ap- 

 proximating volatile substances, such as NHg, alcohol, etc., i.e. 

 by disturbance of the tension, without the membrane bursting. 

 In this process, in spite of the thinness of the membrane, 

 currents must pass one over the other, afferent and efferent 

 currents, otherwise the thin membrane would burst at once. If, 

 however, it was a question of an oil membrane, such as Quincke 

 assumes, it would not be possible at all for it to burst under the 

 given conditions, since its tension at the surface bounding it 

 towards the surrounding water must in any case be considerably 

 greater than that'at the surface which limits it towards the proto- 

 plasm, for which reason it could not be made to burst from this 

 side. I must therefore regard the double currents in a thin 

 fluid membrane of this kind as being quite within the bounds 

 of possibility. 



For the rest, some advance may probably be made at this 

 state of the question by means of suitable experiments. 



(A) Independent Movements of Granules 



Certain points in the problem of protoplasmic move- 

 ments must, however, I think, be briefly touched upon now. 

 Amongst them the question of the relations of the so-called 

 protoplasmic granules to the streaming movements seems to 

 me one of the most fundamental. It is ;notorious that the 

 movements of protoplasm have frequently been estimated 

 only by the movements of these granules which have 

 served to a certain extent as indices of the streaming move- 

 ments of the protoplasmic ground substance. In keep- 

 ing with this there exists an almost universally occurring 



