392 i BOTANICAL GAZETTE [NOVEMBER 
G. 28. Longitudinal section at right angles to “a 27 passing through the 
serie cia ridge (zz) near the points of the crescen 
Fic. 29. Cross sections of embryo at about same age as figs. 27 and 28. 
Fic. 30. Section d of fig. 29; cl, cotyledonary lobes; other letters same as 
in figs. 27 and 28. 
Fic. 31. Reconstruction from sections of embryo a little older than those 
in figs. 27-30. 
Fics. 32-40. Cabomba ED 
Fic. 32. Two-celled embryo. 
G. 33a. Section of embryo in octant stage showing two-celled suspensor. 
Fi G. 336. Endosperm of 33a. 
Fic. 34. Spherical embryo with three-celled suspensor. 
G. 35. Spherical embryo showing original two-celled suspensor subdivided 
by os divisions. 
1G. 36. Longitudinal section of embryo at about time of origin of cotyle- 
donary ridge, showing two-celled suspensor; cut same direction and showing 
same points as fig. 27. 
Fic. 37. Reconstruction from section of embryo slightly older than fig. 36 
and showing early development of cotyledonary lobes. 
1G. 38. Reconstruction from sections of almost mature embryo; drawn to’ 
same scale as fig. 37. 
Fic. 39. Cross-sections of embryo; 2d, 6th, 8th, 12th, 24th, and 28th sections. 
Fic. 40. Endosperm from embryo near same age or little older than jig. 38. 
