N. E. BROWN OX THE STRUCTURE OF DIATOMS. 335 



closing these cavities were solid, imperforated films of silex ; no 

 osmotic theory will account for it. 



Also it is quite certain that there is some extrusion of motile 

 living matter from the interior to the exterior of the diatom, 

 which is controlUd by the will of the organism. 



No one has yet been able to detect any protoplasmic filaments 

 or pseudopodia (other than the crest of protoplasm along the 

 raphe) protruding from the pores of diatoms, and if they are as 

 fine as the pores I have seen would seem to indicate, and as trans- 

 parent as protoplasm, I doubt if we ever shall see them on the 

 living diatom, as the nearness of their own refractive index to 

 that of water would not provide sufficient contrast to enable us 

 to detect them. Killing and staining do not seem to prove 

 successful in demonstrating anything of the nature of pseudopodia, 

 only the crest at the raphe and a very thin layer of protoplasm 

 sometimes covering the whole shell can be made evident, 

 so far as I have been able to demonstrate it, but it ought not to 

 be lost sight of that there is a possibility that a diatom may be 

 able to speedily retract any protoplasmic matter that it may 

 protrude from its shell or from the film of protoplasm that some- 

 times covers its shell, so that at the slightest indication of the pre- 

 sence of anything injurious, all external protoplasm of the nature 

 of pseudopodia may be suddenly withdrawn before the diatom is 

 killed. Usually there is no evidence that any living matter is 

 protruded to any distinct distance from the shell, except at the 

 raphe, as any substances taken hold of by a diatom are generally 

 seen in apparent close contact with the shell, although occasionally 

 one is seen dragging a niece of dirt along at a short distance 

 behind it by an invisible thread. But upon a few rare occasions 

 I have witnessed a diatom seize and move pieces of dirt that were 

 at an appreciable distance from the shell, and on one occasion 

 last autumn I was able to measure the interval between the 

 diatom and the dirt. I was observing a large species of Surirella, 

 probably S. biseriata, which was moving rather quickly across the 

 field, when I saw it seize with invisible hands a large piece of dirt 

 at a little distance from it, and pull it along by its side, without 

 decreasing the distance between itself and the dirt. I at once put 

 on an eye-piece with a micrometer scale on it, and carefully noted 

 the distance separating the dirt and diatom upon the scale, and 

 then substituted a stage micrometer for the diatom and found that 



