EXPLANATION OF Till'. MOVEMENTS 313 



as has been said, so especially characterises tin- posterior 



end, is, judging IV. >m tin- course it takes, somewhat as follows. 



As has been remarked above, it is very prohable thai 

 the surface of the body of tin- Anm-ha is formed by a 

 viscid membrane-like layer. The backward How of the 

 pseudopodia, like the gradual retraction of the hinder end 

 of the Aimeba- that moves forwards in a simple manner, 

 depends, in my opinion, only upon the cessation of the 

 forward current, and the flowing bark of the protoplasm 

 towards other directions. J hiring this jinn-ess the pseudo- 

 podium or the hinder end diminishes continually in si/e, 

 and meanwhile the viscid external layer cannot How 

 together rapidly enough to keej> pace with this diminution, 

 in consequence of which it is thrown into folds, whirl i 

 appear as papillose processes. The whole phenomenon -ives 

 just the appearance of a viscid, membrane-like layer, wliidi 

 shrinks into folds on account of the gradual emptying of its 

 contents. "With continued emptying of the contents these 

 folds become more numerous and fine, so that the com- 

 j)letely shrunken layer is finally attached in the form of a 

 tuft-like bundle of fine processes. These processes or folds 

 then become closely ajDposed, and finally stick together. 

 Meanwhile they are removed from the action of the external 

 medium, which, as we suj>j)osed, is the cause of the cuticle - 

 like toughness of the superficial layer ; this removal, as well 

 as the influence of the interior of the Anxeba, which now 

 makes itself felt upon the tuft, has the effect of gradually 

 making it become fluid ; and the folds or processes fuse in 

 the above-described manner, whereupon the jmxluct of tin- 

 fusion and the rest of the protoplasm finally How together. 



That the origin of the tuft is really due to shrinkage 

 of the membrane-like external layer is also jn-obable from 

 the fact that by treating Anml proteus with fairly concen- 

 trated XaCl solution (4 per rent), the same j>apillar 

 processes can at once be produced over the whole surface, 

 accompanied by considerable shrinkage of the Anm-ha. ami 

 in this case the processes certainly arise by folding of tin- 

 external layer, as a result of a decrease in the volume of its 

 contents. With 0'02 j^er cent caustic jiotash I also saw 



