DEVELOPMENT OF THE OSTRICH FERN. 19 



by a septum, and rapidly elongating assumes the form of the first (PI. 4, fig. 5 /■'). Like 

 this it contains little or no chlorophyll. As growth progresses the chlorophyll granules, 

 ■which had at first been crowded together, become more separate and very distinct. The 

 protoplasmic layer in which the chlorophyll granules are embedded lines the walls of the 

 cells, the center of the cell being occupied by a large sap-cavity. The nucleus, which 

 is sharply defined, and usually contains an evident nucleolus, is round in outline, and lies 

 close to the wall of the cell. In the root-hairs the protoplasm, near the growing end, is 

 finely granular, and a few larger granules are scattered through the periphei-al protoplasm. 

 In the young hairs, some of these are f\iintly colored with chlorophyll but this seems to 

 be finally lost. 



If the protoplasm of the cells of the prothallium is contracted, e. g., by a solution of 

 sodium chloride, and then stained, the continuity of the protoplasm can be easily dem- 

 onstrated. Delicate threads of protoplasm can then be detected connecting the proto- 

 plasmic masses of adjacent cells, apparently passing through small openings in the 

 wall. 



By the end of the second week the flattened form of the older prothallia begins to be 

 marked, on account of the division of the cells in two planes, and about a week later 

 the larger ones show the beginning of the sinus in the fi'ont margin, causing it finally 

 to assume the heart-shape usually met with in the larger prothallia of ferns. This form 

 is usually quite pronounced by the end of the first month (PI. 4, fig. 13). It is brought 

 about in the following way: after a row of several cells has been formed, the terminal 

 cell becomes divided by a wall nearly at right angles to the earlier ones. In the cells 

 thus formed, transverse walls are formed which, however, generally diverge somewhat 

 from an exact i-ight angle with the median wall, being turned slightly forward, thus 

 forming two cells that are nearly triangular in shape. One of these becomes the apical 

 cell of the prothallium, the other behaving like the ordinary cells. The apical cell be- 

 comes next divided by a wall parallel to one of its lateral walls, and next by a similar 

 one parallel to the other side. This process is repeated, resulting in two series of seg- 

 ments lying altei'nately right and left, thus rapidly increasing the breadth of the apex 

 of the prothallium (PI. 4, figs. 7-11). Occasionally, two segments appear to be cut off 

 from the side in succession, but this is probably exceptional. The first division in each 

 segment is transverse, and divides it into an inner and an outer cell. The succeeding di- 

 visions do not alwaj's follow the same order, but usually the next wall divides the outer 

 cell into two equal parts, and is perpendicular to the first one. Following this, the inner 

 cell is divided into two by a ti'ansverse wall. Beyond this point there is great irregu- 

 larity. However, the longitudinal growth of the segment exceeds the lateral growth, 

 so that the end of the segment is pushed out beyond the apical cell. This combined 

 with the rapid division of the marginal cells of the segment results in the formation of 

 a deep sinus or cleft, in the front of the prothallium, in the bottom of which lies the 

 apical cell. "Where this growth is excessive, the two lobes of the prothallium sometimes 

 overlap. Sooner or later the apical cell ceases to form lateral segments, and becomes 

 divided by a transverse wall into an inner and an outer cell, in the same way that the seg- 

 ments were first divided (PI. 4, fig. 12). The outer cell divides into two equal cells 

 by a wall at right angles to the first, and the apical cell thus becomes obliterated, acting 

 from this time on like the other cells. In cases where the prothallium is branched, as 



