182 THE A3IER1CAN MONTHLY [June 



the faculties required iu minute examination — the Pleu- 

 rosigma angiilatuuj — whicli has iu itself a history singu- 

 lar iu the various waves of opinion and attempted dem- 

 onstration. The markings of its silicious envelope at 

 first presented striaB, which further magnification deter- 

 mined into a series of semi-circular bosses, or at other 

 times, according to other views, so many depressions or 

 apertures. The first was once attempted to be illustra- 

 ted by a glass tumbler, the sides of which consisted of 

 so many raised bulbs. It was thought that a similar 

 material would be similarly att'ected by the action of light 

 and thus would prove, or tend to prove, the true con- 

 struction of the valve. In the theory of elevations, it is 

 not so long ago that arrangements were made in side 

 illuminations by a pencil of light, thus supposing to give 

 a true and artistic light and shade. But, in both these 

 nice experiments, it seemed to be forgotten that they 

 were begun in a foregone conclusion ; and, as I have 

 previously said, you naturally, in such a case, see what 

 you wish to see. Certain accidents, fractures, and pecu- 

 liarities inconsistent with the above-named views, assis- 

 ted by careful illumination, seem now to have tolerably 

 settled the question to be on the side of apertures, and 

 my predecessor has worked successfully thereto. That 

 this must be the general consent on such markings 

 throughout the DiatomacesB must probably be entertain- 

 ed, though it would be dangerous to affirm that there 

 was no variation from it in the multiform changes of 

 nature. 



But the subject has been so admirably worked out and 

 recorded by two papers in our Journal, one by Mr. C. 

 Haughton Gill, April, 1890, in which is well described 

 his mode of preparation of the objects wherewith to 

 determine the structure. Another by Mr. Nelson, in 

 May, 1891, goes into the same matter by the use of high 

 powers, and these papers, showing a working on differ- 



