158 PALMKk : 



siderable portion of the keel is applied to the glass. Then it 

 begins moving. Sometimes it will go straight along, hanging 

 by one keel only, as in the cross-section diagram 5. It is 

 accustomed to swing more or less as indicated by the arrows, 

 somewhat as does the Ni/asc/iia. At other times it will swing 

 so far that two raphes are applied to the glass at the same 

 time. As a rule, it hangs dangling down as in the figure, 

 with a slight tendency to roll sidewise just as when upon a 

 substratum — in both cases following the curve of the keel. 

 All these gyrations seem to me best explained by the view 

 that they are due to a minute ribbon of contractile protoplasm 

 reaching through the narrow raphe in the keel, adhering to 

 the glass, and moving back and forth along the length of the 

 raphe. In figure 4, on this view, there would be a firm but 

 very short contact at D. The protoplasm forming this con- 

 tact begins to move along the raphe in the direction of E, 

 retaining its grip upon the glass. The result would be a 

 smooth sliding of the frustule up(jn its keel in the opposite 

 direction, until the lower end gradually rises and a longer 

 section of the raphe comes into reach of the glass. Then, if 

 the current continues, motion of translation along the glass 

 will be the result, and this motion will continue so long as the 

 current endures, and the grip of any portion of the ])rotrud- 

 ing plasma remains good. The transverse swaying motion is 

 also capable of explanation on the same lines, but not fif elu- 

 cidation by diagrams on a flat surface. 



Having vouched for the reality and the substantial accu- 

 racy of these facts, I would once more remark upon the impos- 

 sible character of any general theory of diatom motion based 

 on inward or outward streams of gas. water or i^allertc, and 

 express a renewed antl greatly recnforced conviction that 

 moving protoplasm is the real and direct activator. Cling- 

 ing and motion under such precarious conditions are of the 

 very type of the wiertl doings of protoplasm. We see the like 

 in the rhizopods ])articnl:irly, where the possession of the 

 power to leave the slide and tlraw up to and crawl on the 



