504 



THE FORMS OF TISSUES 



[CH. 



the focus or axis of a cell; if they be hghter they drift to the 

 boundaries, then towards the nodes, where they tend to form 

 tri-radiate figures like so many 'Hri-radiate spicules". But if they 

 be in very fine suspension a curious thing happens: for as they are 

 carried round in the vortex, the lowermost layer of hquid, next to 

 the solid floor, keeps free of particles; and this "dust-free coat*", 

 rising in the axis of the cell and descending at its boundary-walls, 

 surrounds an inner vortex to which the suspended particles are 

 confined. The cell-contents have, so to speak, become differentiated 

 into an "ectoplasm" and an "endoplasm"; and an analogy appears 



A B 



Fig. 182. Benard patterns in smoke : ^, at rest; B, under shear. After K. Chandra. 



with the phenomenon of protoplasmic "rotation," where the outer 

 layer of a cell tends to be free from granules. When bright glittering 

 particles are used for the suspension (such as graphite or butterfly- 

 scales) beautiful optical effects are obtained, deep shadows marking 

 the outlines and the centres of the cells. Lastly, and this is by no 

 means the least curious part of the phenomenon, the free surface 

 of the Hquid is not plane; but each little cell is found to be dimpled 

 in the centre and raised at the edges, in a surface of very complex 

 curvature!, and there is a curious pulsation in the flow, especially 

 when waxes' are used. 



* Cf. Tyndall, Proc. Roy. Inst, vi, p. 3, 1870. 



t The differences of level are of a very small order of magnitude, say 1 /x in 

 a layer of spermaceti 1 mm. thick. 



