112 ANDREWS. 



more marked and refractive as they near the edge, where they 

 are so closely set together as to appear fused into a thick, 

 highly refractive rim. At this time the collar has perceptible 

 and constant viscidity, amounting to rigidity, for it bends not 

 at all to sharp pressure from other larger organisms, as from 

 the appendages of Stylonichia moving about it. 



From the standpoint of a physical and mechanical hypothesis, 

 what marvellous and irrevocable transformations of substance 

 must instantaneously have taken place in response to that 

 passing jar of the cover glass! 



Yet after a few moments' quiet the striae begin to be less 

 distinct and the shape of the collar shows a new change. It 

 opens now at the top, widening gradually until it again equals 

 its first expansion. As it nears this limit, another set of striae, 

 but far more delicate, appears encircling the collar, that is, 

 transversely to the first set which has by then quite disappeared. 

 Finally all striation vanishes, and by their renewed movement 

 up and down over the surface of the film the pigment granules 

 show that the former fluid state is again active. That it is truly 

 a fluid state is shown by the whole film being at other moments 

 returned by flow into the body and then again extruded in any 

 one of a number of fantastic ways. It may appear next as an 

 amoeboid mass extending lobose processes, or beyond these filose 

 threads, or as a huge bubble of protoplasm such as Vorticellidae 

 are found blowing, the walls of which may here be thick or 

 thin, smooth or variably lumpy. This will thin out more and 

 more at the uppermost convexity until it seems to burst like a 

 bubble. After this the whole expands itself gradually, — the 

 irregular protoplasm returning into the body, — and the normal 

 appearance of the "collar " is restored. The flagellum maybe 

 reproduced from the centre of the collar included area by an 

 irregular pseudopodial process which undergoes swift recon- 

 struction until the normal flagellum is again actively functioning. 



The collar will again later suffer metamorphosis into some 

 new mask for the substance. How unimportant here does the 

 form of the substance seem in comparison with its habit! 



If a cross section of a collar film could be made, equalling in 

 depth the thickness of the film, we should have a filose-like 



