1 902 J ON THE DEVELOPMENT OE CERTAIN PIPERACEAE 329 



tiated except for a slight projection above, which perhaps is a 

 secondarily developed, rudimentary suspensor {^fig. 2g). As 

 the embryo pushes down into the endosperm the latter seems to 

 close in above so that the embryo often comes to lie nearly 

 centrally in the endosperm. 



The germination of the seed in Heckeria iimbellata {fgs, 30,^1) 

 is essentially like that of Peperomia pellticida, and will be men- 

 tioned in connection with the discussion of this process in the 

 latter genus. 



III. THE GERMINATION OF THE SEEDS OF FEPEROMIA AND 



HECKERIA. 



The ripe fruit of Peperomia pellticida is oval in form^ with a 

 slightly pointed upper end, being about O-Ss"""" long and 0.6'""^ 

 wide. It is made up, beside the seed, of four or five layers of 

 carpellary tissue. The outer or epidermal layer is made up of 

 bulging, thin-walled cells interspersed with knob-like glandular 

 hairs, perhaps hydathodes in function {fig-J^^) The inner layer 

 of the ovary-wall is of very large cells, with coarsely reticulate 

 outer walls and with the basal walls irregularly thickened with 

 granular material {figs. J2, jj). Between this layer and the 

 epidermis are two or three layers of delicate, flattened cells. 



The seed proper is 0.6"'"^ l^^g ^^d 0.45"^"^ in transverse diam- 

 eter. It is surrounded by a single integument, the two cell-laj-ers 

 of which have their walls thickened and of a very dark color. 

 The inner walls of the outer layer of the integument fit against 

 the cells of the perisperm, especially at the upper end of the 

 seed, with a very wavy outline {figs. J2, jj). In an earlier 

 paper (Johnson, igoo^, fig. 75) I was misled, by the much more 

 regular line of cell-walls separating the outer layer of perisperm 

 cells from those within, into thinking this the line of separation 

 between integument and perisperm. The interpretation given 

 above is however the correct one, as has been shown b}^ Camp- 

 bell (1901, p. 3), and the lower index line from m/ in the figure 

 IS referred to above, should run only to the inner, wavy border 

 of the inner of the two darkly shaded cell-layers. All of the cells 

 represented as containing starch in this figure belong to the peri- 

 sperm. The great bulk of the seed within the integuments is 



