DIATOM LOCOMOTION. I57 



therefore less close than that described for A^. Iridis. But 

 alike when the groups are flat and when they are as here indi- 

 cated, the diatoms have no special trouble in retaining their 

 dependent position, and they move about under the cover 

 glass approximately as well as when upon the slide. When, 

 however, the last raphe is by any means caused to part com- 

 pany with the glass, the whole group falls down through the 

 water. The tenacity with which these diatoms cling may 

 be understood on considering this diagram of a group of 

 four held in place by only one of the number. No facts of 

 interest were gathered from behavior in ink emulsion. 



Nitzschia linearis. This is roughly and inaccurately repre- 

 sented in cross section at 3. The small keels of this diatom 

 are curved in reverse directions, and each ends in a very thin 

 edge, where runs a raphe or cleft from end to end. Being 

 very long, very narrow, and unusually flat besides, one is left 

 to wonder how it manages to crawl so vigorously and so con- 

 tinuously on a flat surface, if it be necessary that a raphe 

 shall always be in contact with the substratum. However 

 this may be, A^ linearis is one of the most able of all at 

 crawling when suspended by one raphe from the under side of 

 the cover glass. As it goes, it sways from side to side, 

 swinging as on a hinge, in the directions indicated by the 

 arrows. Hanging clear, held only by this knife-edged keel, 

 it goes ahead rapidly and jostles loose other diatoms larger 

 than itself. There is not a sign of anything to explain the 

 basis and means of these amazing activities, which meantime 

 continue with every aspect of a complete competence for an 

 indefinite number of repetitions. 



Surirella li?iearis and S. splendida. Both these species 

 suspend themselves upon the under side of the cover glass, 

 and move about very freely. The schematic figure .\ repre- 

 sents a longitudinal view of a Surirella in an extreme posi- 

 tion, with only one end of a keel in contact. It does no 

 crawling in this attitude, but by some sleight it will raise the 

 lower end gradually in the direction of the arrow until a con- 



