466 Transactions of the Society. 



minute and delicate objects, and who agree that all real increase of 

 our knowledge, even in these branches, has been originally obtained, 

 or at least could have been as well obtained, by powers not much 

 exceeding 1000-1200, whatever kind of lenses may be in question. 



2. I cannot, however, restrict myself to the suggestion that 

 excess of power, in proportion to the aperture in use, is simply of no 

 advantage for the recognition of microscopical objects, but I must go 

 further still, and express the opinion that excessive power is, or at 

 least may be, a positive obstacle to correct recognition, because it 

 will unavoidably lead the observer to take mere optical pheno- 

 mena of the images for real attributes of the objects. The following 

 considerations will justify this view. 



If we observe a frustule of PI. angulatum with the small aper- 

 ture of about • 6, in which case one set of lines only is exhibited at 

 once, we may obtain a well-defined image of these lines under a power 

 of 1000 diameters, or even more, provided a relatively short focal 

 length of the objective and a moderate eye-piece are applied, and 

 the illumination is effected by a very narrow beam of intense light. 

 This power apparently displays much more than a lower power of 

 350 or 400 diameters with the same aperture. We see the strige 

 as broad ridges or grooves widely separated from one another, and 

 we recognize a distinct proportion between the breadth of these 

 apparent ridges and their interspaces, which is 1 : 1 very approxi- 

 mately ; whilst with 300 diameters we only catch just the fact of 

 a striation, and nothing more. In this case now, we know that all 

 details which are given by the 1000 diameters are mere optical 

 illusions, because we are able to control and correct the indica- 

 tions furnished by that power with the low aperture, by the 

 image presented by an equal power, but having an aperture of 

 1 • 2. But if it had happened that a system of wider aperture 

 than 0*6 had not hitherto been made, microscopists would cer- 

 tainly have believed in the existence of ridges and grooves of 

 equal breadth on the scale of PI. angulatum. In this example it 

 is unquestionable that the image obtained with an aperture of • 6 

 under 300 diameters is less far from the truth than the image 

 with the same aperture under 1000 diameters. The indeterminate 

 striation is an indication of real structure, inasmuch as there are 

 equidistant rows of elements in the diatom, which must appear as 

 striae as long as the elements themselves remain occult ; the 

 exhibition of these rows as determinate ridges and interspaces with 

 a distinct relation of breadth is a positive adulteration of the image 

 of the structure. 



What holds good for an aperture of ' 6 must also hold good 

 for every larger aperture relatively. I have before me Dr. Wood- 

 ward's magnificent photographs oiAmphipleura joellucida, PI. angu- 



