503 



every case by a pellicle, this pellicle must also invest the outgrowing 

 process, and form the delicate wall of a tube filled by cell-plasm. 



The tube which is thus formed may be cylindrical or may be 

 flattened — whe should expect the latter form in cilia which are to 

 work in one plane only — but for the most part their minute size 

 prevents an exact determination of this point. And that a further 

 differentiation takes place in the minute tubular outgrowths which are 

 originally formed is clear from the observation of Franze in Poly tomes 

 to the importance of which I have already drawn special attention. 

 The conclusion therefore is that a cilium is a tubular outgrowth from 

 a cell occupied by protoplasm (probably by liyaloplasm) and bounded 

 by a pellicle or cuticula which in most cases has undergone differ- 

 entiation in such a fashion as to cause the movements of the cilium 

 to be executed in a definite manner or in a single plane. 



But before this differentiation is complete we should expect the 

 protoplasmic outgrowth to retain many of the functional manifestations 

 of undifferentiated cell -protoplasm. The most prominent of these 

 manifestations is met with in the form of local changes at the surface 

 of the cell; produced, there can be no reasonable doubt, by variations 

 in tension of the superficial layer, and these variations in surface ten- 

 sion must be due to local chemical changes caused by agencies pro- 

 ceeding either from within (spontaneous changes) or from without 

 (stimuli). And equally in the phylogenetic series we should expect to 

 find in the lowest terms of the series (Protista) modifications of ciliary 

 action which would tend to show that their cilia have undergone less 

 differentiation than those of Metazoa, less regularity and more inde- 

 pendence of action, as is indeed abundantly illustrated in several of 

 the passages which I have quoted. It is in this way that the move- 

 ments of entirely isolated cilia are to be explained; for, with the ex- 

 ception of the single observation of Meves upon the severed cilium of 

 the spermatozoon of the Salamander, all the observations of movement 

 of entirely isolated cilia which have been recorded are from Protozoa : 

 in the case of Metazoa it would seem to be essential that some part 

 of the cell-protoplasm should remain in connection with the cilia. The 

 conclusion therefore is that it is only the less differentiated cilia that 

 are capable of exhibiting variable and independent movements, retaining 

 to a considerable degree the property of undergoing local changes of 

 surface tension which is common to all undifferentiated protoplasm. 



But it may be asked: Given such tubular structures enclosed by 

 an undifferentiated pellicle as we have reason to believe these inde- 

 pendently acting cilia to be, how do we know that local changes of 



