N. ROSÉN, STUDIES ON THE PLECTOGNATHS. 15 



narrower and pointed. Whether the bundles whicb run paral- 

 lelly to the sides of the hexagon end abruptly at the side of 

 the triangle, or perhaps, send some branches into the »lines», 

 I cannot tell. Vertical bundles are quite wanting in the 

 lines. The lower surface of the triangles, which is not hori- 

 zontal, ; but sloping from the interiör angle downward towards 

 the side of the hexagon, is lined by a thin stratum of fibres. 

 This layer is in continuation with the »lines» and is formed 

 in an analogous manner. The bundles in the same direction 

 as those forming the lines but running into the deeper por- 

 tion of the triangle continue af ter leaving the triangle for 

 some distance, as shown in Fig. E. The centre of the plate 

 is beneath the homogeneous layer consisting of horizontal 

 bundles of fibres crossing each other in several directions. 

 The fibrillated part of the plate hitherto described can be 

 compared with a six-sided pyramid with thin walls and ex- 

 cavated below (= the stratum below the triangles). The top 

 is cut off. The radiating edges of the pyramid form narrow 

 ridges, the »lines». The spaces between the ridges are filled 

 up with the »triangles». There seems to be another difference 

 between the triangles on one side and the »ridges» and the 

 surfaces of the pyramid on the other side, than that caused 

 by the different arrangement of the bundles of fibres. From 

 the examination of non-decalcified plates it is evident that 

 the parts described as ridges and surfaces of the pyramid 

 are harder than the rests of the fibrillated portion of the 

 plate. A section through these parts shows a more shiny 

 surface than through the »triangles». The chemical diffe- 

 rence I have not analysed. It is of no interest from a mor- 

 phological point of view. Sections of non-decalcified plates 

 stained with van Gieson's mixture tend to proove a diffe- 

 rent chemical nature of the above mentioned parts. The 

 whole plate has not yet been described. Taking up the 

 comparison with a hollow pyramid, we can best describe 

 the rest of the plate by saying that the interiör of the pyra- 

 mid is filled up by a great number of horizontal lamellae. 

 Each lamella consists of fibres running in the same direction. 

 Four different directions alternate with each other. Fig. E 

 shows this part diagramatically. The plates are, as already 

 Williamson has found, penetrated by canals. It is especi- 

 ally the homogeneous portion of the plate that is richly 



