1900] LIFE HISTORY OF SILPHIUM By 
corresponds to the cell similarly placed in Capsella; for in 
Taraxacum this cell produces only concentric layers of derma- 
togen and periblem, a second suspensor cell doing the same 
thing, while the root cap is formed by the third cell from the 
embryo cell. In such a case, the term hypophysis, as applied 
to a single cell, could be justified only by proving that these 
three cells arose by the division of an original single cell next 
to the embryo cell. This, however, has never been done. 
As a further preface to the discussion of the development of 
the embryo, it may be well to state that in the early part of the 
work the entire ovary was cut longitudinally in the plane of the 
radius of the inflorescence, in order to determine with certainty 
the direction of the early divisions in the embryo. Later, the 
ovules were dissected out, and the direction of the section, 
whether in the radial or tangential plane, was ascertained from 
the position of the funiculus. 
Passing on now to the second division in the young embryo, 
we see by fig. 68 that the two nuclei present may prepare for 
division at almost exactly the same time. The nucleus of the 
terminal cell is slightly in advance of the other, as the segmenta- 
tion into chromosomes has occurred, and the nuclear membrane 
has disappeared. However, the difference is so slight that we 
are left without absolute proof as to the origin of such a stage 
as fig. 69. If the terminal cell alone has divided, two interpre- 
tations are possible. First, a transverse wall may have arisen in 
the embryo cell, in which case my jigs. 69, 70, 71 would cor- 
respond exactly to Schwere’s figs. 8, 9,a, 9, respectively,as fig. 70 
is cut in the radial plane and fig. 77 in the tangential. The 
second possibility is that the terminal cell in fig. 6g is the real 
embryo cell, but marked out by the second division in the 
spore ; and a comparison of figs. 69, 70 with 72-76 shows that 
this is actually the cell which produces the octants. This being 
true, we may now apply the term embryo cell definitely to 
the terminal one of the three cells in fig. 69. It would be an 
obvious error to give.that name to the terminal cell of fig. 67, . 
for, according to this interpretation, the latter cell produces 
