36£ Dr. G. C. Wallich on the Structure of 



In five divisions of the micrometer, diagonal lines 10 

 u „ longitudinal do. 15 



u „ „ transverse do. 12£ 



In a space equal to ten of the dia- 

 gonal lines on the diagram, longitudinal lines 14f 

 „ „ „ transverse do. 13 



It is perhaps necessary to add that, in speaking of lines, I 

 mean to convey the idea of perfectly clear and well-defined lines, 

 such as are visible on the finest kind of cut glass or crystal by 

 the naked eye — and further, that, with the highest powers, it 

 becomes perfectly manifest that the diagonal series which bound 

 the elevated portions of the structure are not "bent at short 

 intervals at an angle of 120°," as supposed by Schultze and 

 others. On this head I may remark that if the sides of an 

 hexagonal figure, such as has been assumed to exist in P. angu- 

 latum, are resolvable, inasmuch as each of the sides must, a 

 priori, occupy the length of one of these "bends," it follows as 

 a necessary consequence that the same degree of magnifying 

 power, and the same adjustments of the microscope, ought to 

 render apparent the " bends " or zigzag formed by the succes- 

 sion of these deviations from a direct course. But inasmuch as 

 these powers and adjustments avowedly fail to do so, it seems 

 almost like an assumption unsupported by evidence of any direct 

 kind whatever to regard the lines otherwise than as they appear. 



Again, assuming the lines to be resolvable into " dots " or 

 hexagons, and accepting the estimate laid down in the ' Synopsis 

 of British Diatomacese' as correct, namely, 52 in '001", and 

 further assuming the diameter of each of the hexagons to be 

 double the diameter of the " bent lines " by which they are 

 bounded (which is about the proportion deducible from Mr. 

 Wenham's photographic representation), it is manifest that the 

 diameter of the "bent lines" would be about one 100,000th of 

 an inch, and that this would also be an approximate estimate of 

 the length of each deviation to the right or left of the direct 

 line. Now, since it is well known that lines of this degree of 

 thickness can be clearly defined, and even counted, when not 

 closely aggregated together, it follows that the " bent " form 

 ought to be, and indeed would be, definable, did it really exist. 

 But, as candidly acknowledged by Schultze, " these bends are 

 imperceptible " with powers even of 500 and up to 800 dia- 

 meters ; and this being the case, the inference is surely war- 

 wanted, that no faith can be placed in the apparent outline of 

 the spaces said to be actually determined and defined by those 

 " bent lines." 



But, as I have formerly mentioned, in the Pleurosigmata there 



