OF CASTALIA ODORATA AND NyMPHAEA ADVENA 217 
the cotyledon growth and eventually forms the suspensor which 
projects towards the micropyle as a short cone-like organ, but does 
not become attached to the walls of the embryo-sac (/. 29, 5, 
32, S) and does not appear to be of any great value. 
The cotyledon was not observed to bifurcate (¢ext fig. 2), as 
Lyon describes for Ne/umbo (g), although embryos older than 
rig : h i 
Fis. 2. Outline drawings from series of sections of single embryo of WV. advena 
showing relation of parts. 5, suspensor; #, plumule; #, cotyledon. 
those figured were examined. Soon after the appearance of the 
Suspensor there is an increased activity of the remaining spherical 
Part of the embryo, resulting in the formation of a plumule (/ 
31, plum.). This agrees with Ne/umbo (g) as described by Lyon. 
he primary root is formed by an increased activity of cells be- 
tween the suspensor and plumule (f 37, 7) and also agrees with 
Nelumbo as described by Lyon. 
The embryo in its development cannot be said to show a 
marked resemblance to any of the monocotyledonous plants to 
| which we might suppose it to be related. Aside from the suspen- 
Sor it compares fairly well with Lilium and Erythrontum. How- 
