112 THE AMERICAlSr MONTHLY [July 



learning what the actual detail is," because no objective 

 will embrace the other orders. Examining these several 

 statements, there is every reason to believe that a dry 

 objective with a wide cone oi light gives a perfectly 

 truthful image, while it will give the hexagons quite 

 easily if that figure is preferred ; Zeiss's well-known 

 large-scale photograph is of a value so coarse that it is 

 beyond dispute that a portion of the second-order spectra 

 were included by the lens used, with the result of intro- 

 ducing a false doubled resolution impossible with first 

 orders alone; and an immense further step can be taken 

 by using a first-rate immersion-lens of 1-40 aperture, 

 with a wide cone. The Zeiss photograph X4900, and 

 the Van Heurck photograph, are confessedly the highest 

 triumphs of photography by the Abbe method : one has 

 only to compare both with the beautiful photograph 

 X4900 taken in this other way by Mr. E. M. Nelson, and 

 other similar ones up to a scale of X 6400, to see once for 

 all, which is the truest image, and the all-importance of a 

 suflBciency of heterogeneous light. 



The minute detail in some of these photographs could 

 not possibly be shown by that method, because, minute 

 as they are, they are unsymetrical and not periodic. In 

 regard to the P. angtilatum, both circular disks and hexa- 

 gons can be seen, depending upon the precise focus ; the 

 sharpest portions show the circles, which, disposed in 

 quincunx arrangement, most diatomists who have 

 worked with English appliances believe to be the true 

 figure. Besides the sharpest image, we have the pheno- 

 mena of "postage-stamp fracture," and the shape of far 

 coarser markings in other diatoms to guide us. Mr. C. 

 Haughton Grill has demonstrated that the spots are 

 either apertures or depressions, by depositing pigment 

 in them ; and the various images can be imitated with 

 perforated zinc. It is the distinct outlines of the 

 fractures, and broken-through apertures, which are so 



