1892.] NEW-YORK MICROSCOPICAL SOCIETY. Ill 



portions of the cell wail gives the appearance of beads strung on a 

 stout wire. Fig. i r shows a few of the polygonal cells of the 

 brown substance of the cleft, taken from the last portion of the 

 transverse section at ^, Fig. 7. 



The longitudinal-radial section so closely resembles the trans- 

 verse section that it would be very difficult to distinguish them 

 if the labels were removed from the respective slides. Therefore 

 the description of one will answer well for the other, showing that 

 the cells of the inner structure are successively prisms, spindles, 

 and irregular globular cavities. 



The longitudinal-tangential section presents quite a different 

 appearance. Fig. 12 shows a small portion of such section, taken 

 quite near the surface of the stone, but after several successive 

 slices have been removed. The left-hand portion of the figure 

 represents the outermost layers of the seed coats. First are found 

 long, fusiform, hyaline cells, presenting no marked structure with 

 high magnification. Then succeed still longer, attenuated cells, 

 with much thickened cell walls, minute lumen, and exceedingly 

 abundant capillary canals running from the lumen to the middle 

 lamella, there meeting similar canals from adjoiningcells. Fig. 13 

 gives an enlarged view of three such canal cells. These cells and 

 the succeeding cells of the seed coats do not always lie parallel 

 with the longitudinal axis of the stone. Usually they are longi- 

 tudinal, but frequently they are disposed in groups, which branch 

 at various angles, until sometimes they are directly transverse, 

 parallel with the smaller circumference of the stone. Next with- 

 in the canal cells follow large, irregularly elliptical, dark-brown 

 cells, with slightly pitted walls, and usually disposed in two layers. 

 Then follow more dense tissues, of short, fusiform, hyaline, and 

 light-brown polygonal cells, until we come to the outer ends of 

 the palisade cells of the endosperm, and, as the section dips a 

 little deeper, the outer extremities of the spindles are seen as 

 little circles, of various sizes according to the position in which 

 they happen to be cut, and all imbedded in the dense structure of 

 the thickened cell walls. Fig. 14 gives an enlarged view of the cut 

 outer ends of these palisade cells and spindle cells, as they appear 

 in this longitudinal-tangential section. 



The hard structure furnished by the thickened walls of this en- 

 dosperm is said to be pure cellulose. The interesting fact alone 



