22 GRACE E. COOLEY ON 



formefl, <1h' walls arc very tliiii, and the jirotoplasm streams in delicate threads from 

 the larye nuclei to the walls. The protoplasm, in the masses, which are in contact 

 with the walls, is very granular and contains many tiny oil globules. 



The cells in general contain large quantities of oil and much sugar. 



PI. 5, fig. 2, is a sketch of <a few of the endosperm cells cut in the plane of 

 their long axes; the oil has been removed by absolute alcohol ; at n, is the nucleus. 



The layer of small, close-set cells of the integument, which contain the greenish 

 oil drops, is illustrated in PI. 5, fig. 3. 



The oil in these cells reacts differently toward osmic acid from that of the 

 endosperm, the resulting color with this reagent being deep red-brown, not the 

 ordinary black-brown. 



Alisolute alcohol dissolves it completely but more slowly than it does that of the 

 endosperm. Much tannin is present in the integuments, as is shown by the i-eactions 

 with feri'ic chloride, potassium bichromate, and osmic acid. 



Sugar is also present, and at the chalaza a little starch occurs, outside the endo- 

 sperm. 



On July 10, ten days later, the thickening of the walls had begun in parts of 

 the seed which were near the chalaza, and extended toward the center of the seed. 

 The corners of the cells are first affected, and later the rest of the walls, these being 

 thickened on all sides at the same time, and uniformly. 



The cells contain much oil and i^ranular matters and some suscar, thou2;h the 

 reduction of Fehling's solution does not give so great a precipitate as in the earlier 

 stages. PI. 5, fig. 4, shows cells of the endosperm at the stage when the thickenings 

 have only just appeared at the corners; the section was examined in absolute alco- 

 hol, and most of the oil had been removed. In the outer layers of the endosperm, 

 where the walls are not thickened, the protein matters about the nucleus and the 

 nucleus itself are stained red-brown in the presence of iodine. In these cells are 

 peculiar crescent-shaped bodies, which also stain red-brown with iodine ; the nature of 

 these bodies is not yet cleai'. PI. 5, fig. 5, is a sketch of cells where they occur ; at 

 n, the nucleus ; the section has been in absolute alcohol. 



When the reserve cellulose appears in the endosperm, it is clear and colorless, 

 and is not to be distinguished from the original wall until water is applied, then it 

 begins to swell and, immediately, through the absorption of water, goes over into a 

 jelly-like, mucilaginous modification, and the cell contents are pressed together into 

 the middle of the cell. This process of swelling in the reserve cellulose can be 

 easily followed, when to sections of the seed, which have been in absolute alcohol, 

 water is added, for it is then, of course, much slower. The swelling at the corners 



