Hoyal Microseoincal Society. 3 



Tliese facts being clearly established, it is evident that on the 

 examination of any given species of diatom, or other object, in air, 

 and in bisulphide of carbon (or phosphorus), they are seen under 

 conditions in which the respective optical effects arising from their 

 form, are reversed — -under the first condition (in air), each part 

 having a convex surface, gives a positive image when the body of 

 the microscope is raised, and each concave surface a negative 

 image ; on the other hand, in bisulphide of carbon the reverse of 

 this takes place; the concave gives a positive, and the convex a 

 negative image, because in the former case the concavity in the 

 silex being occupied by the bisulphide, a convex lens of that iiffh. 

 medium is produced, and similarly in the latter case the convex 

 silex forms a concave lens of the same highly refractive fluid. 



There is, however, another condition under which a positive 

 image may be produced — a beam of light passing through a small 

 aperture will form a picture of an object placed beneath, as may 

 be seen in Polycistinae and other microscopic objects, and it is 

 therefore possible that an image may be depicted above a diatom, 

 whether it be mounted in the dry state or in Canada balsam, and 

 it is on this account that the necessity of two media exists. 



With a view to giving some practical effect to these considera- 

 tions I determined on examining some coarsely-marked diatom, and 

 selected for the purpose Coscinodiscus occulus iridis, but before 

 describing the appearances presented I may mention that the bisul- 

 phide of carbon reduces the available aperture of dry objectives to 

 73^ 9', a solution of phosphorus in bisulphide (giving a refractive 

 index of 2) reduces it to GO"^, and pure phosphorus to 52° 40'. 



The first point which attracted my attention on examining the 

 bisulphide slide was a valve, the outside uppermost, and broken 

 through, displaying with an i objective under my form of bino- 

 cular, with perfect stereoscopic effect, the second or inner layer with 

 great distinctness, demonstrating, if that were necessary, that as 

 each diatom consists of two valves, so each valve consists of two 

 layers, making, in the complete fi'ustule, four layers at hast. 



Mr. George Shadbolt* proved in 1849 that the Arachnoidiscus 

 consists of four discoid portions, and of " two annular valves " ; and 

 in a paper by Mr. Charles Stodder, read before the Boston Society 

 of Natural History, in the year 1862,t that writer says he has 

 found a specimen of Coscinodiscus beyond the broken edges of which 

 " was another part of the disc, which was simply granular, with a 

 milky aspect"; and further on he speaks of this as the "inner plate," 

 but adds that it " is composed of spherical granules of silex, joined 

 or cemented together by a thin plate of silex." 



My sole object being to demonstrate, as far as possible, the 



* See 'Transactions of the Microscopical Society,' vol. iii., 1852. 

 t See 'Quarterly Journal of Microscopical Science,' vol. xi.. p. 214. 



B 2 



