36 THE AMERICAN MONTHLY [Feb. 



observed with ease in verifying what is said herein in 

 reference to its intricate motile powers. 



In further studies of Navicula nobilis and firma, with 

 the view to verifying the results obtained by a former 

 contributor on the subject of the movements of diatoms, 

 I made use of methyl-blue to differentiate the epidermal 

 covering or mantle, by the following methods. From a 

 rich gathering of living Navicula nobilis, firma and 

 Surirella biserriata, and other forms, I transferred a 

 drop to a slip, and observed them with a i inch objective. 

 By this means I was enabled to note that as many as 

 twenty forms of N. nobilis and firma could be found in 

 parallel contact at one and the same time, gliding back 

 and forth in contact with each other, somewhat after the 

 manner that colonies of Bacillaria paradoxa move at times. 

 While having this special gathering under study in order 

 to note the character of the epidermal envelope, it be- 

 came relatively easy to note the amount of separation be- 

 tween two or more touching frustules of the surrounding 

 transparent layers. Now, admitting that the external 

 layer, if it exists at all, must have the character of an al- 

 buminous substance, such as the white of an Qgg, the 

 substance ought to coagulate under a boiling temperature, 

 and take on an altered or fixed state the same as the 

 white of an egg does when boiled sufiiciently. By 

 shortly drying such a slide of living diatoms over a stud- 

 ents' lamp flame, and completing the mount with thin 

 balsam, we find that the epidermal covering has been 

 changed to a practically impervious envelope. The thin 

 balsam failed tu penetrate many of the frustules during 

 a period extending over months. The slide on examin- 

 ation perodically showed the frustules to be filled with 

 air, and the shrunken or contracted threads of endoplasm 

 still showed a strong greenish tint in the air filled spaces. 

 On the contrary, it is well know that, in acid-treated dia- 

 toms of like character, there is almost an immediate ex- 



