1644 THE VOYAGE OF H.M.S. CHALLENGER. 



the different positions to wliicli the focus of the microscope is brought, the hexagonal 

 tracery exhibits a different appearance ; either a regular network of equal hexagonal 

 frames, or a lattice with equal and equidistant circular openings, or a combination of 

 three crossed and equidistant systems of parallel lines (crossed at angles of 120°). 

 All the well-known and often discussed optical phenomena which are observed in the 

 valves of the Diatomaceae, reappear also on the shell of the Challengerida. 



To recognise the true nature of this diatomaceous structure, sections and slides of 

 the shell are required, and also fragments of broken shells, the broken margins of which 

 are of special value for obtaining further explanation. The figures given by Dr. J. Murray 

 {loc. cit., PI. A) have already demonstrated that the pit-like depressions are regular small 

 cavities in the flinty shell-wall, which possess an inner and an outer opening. These 

 intraparietal cavities are either spherical (fig. 76) or cylindrical (fig. \d, 4a) or spindle- 

 shaped, with a constriction (fig. 2a). Further accurate examinations probably will show 

 a greater variety in their shape. But each cavity constantly possesses two small openings, 

 one on the outer, and the other on the inner surface of the shell-wall. Closer examina- 

 tion (especially of broken shells and slides) proves the presence of these two openings, 

 which in the smaller species seem to be absent on the first view. I may, therefore, 

 suppose that also in the similar valves of the Diatomacese, which exhibit exactly the 

 same structure, an inner and an outer opening are always present in each hexagonal pit, 

 and that the very small size alone prevents them from lieing recognised. The regular 

 hexagonally-framed cavities in the shell-wall of the Challengerida (probably also of the 

 Diatomacese) are therefore genuine pores, differing from the pores of other Eadiolaria 

 only in their smaller size and the dilatation of the middle part, which is much wider than 

 the two openings. 



The mouth of the shell, its single larger opening, exhibits in the various 

 Challengerida a gi'eat variety in form and structure, and offers the best means for 

 separating genera and species. According to its essential shape two difterent sub- 

 families and six genera may be distinguished in this family. In the first subfamily, 

 Lithogromida, the mouth is a simple large opening in the wall of the shell on the oral 

 pole of its main axis (PI. 99, figs. 1-15). In the second subfamily, Pharyngellida, the 

 mouth is prolonged into a cylindrical hollow tube, which is prominent into the shell- 

 cavity, and comjjarable to the pharynx of the Metazoa in general form and function 

 (PI. 99, figs. 16—20). In this case we may distinguish an inner and an outer mouth 

 of the tubular pharynx. 



In a few forms of both subfamilies the mouth is toothless, smooth, and not armed 

 with prominent external spines or teeth, as in Lithogromia, the simplest form of all 

 (PL 99, fig. 22) and in Entocaymula (figs. 19, 20). In the great majority of 

 Challengerida, however, the outer opening of the mouth is armed with one or more 

 teeth, usually strong prominent spines (figs. 1—18). Usually the base of these teeth is 



