883 



towards the axis of the flower. Dr. Ghaud, on the other hand, as po- 

 sitively asserts that the suspensor passes out at the micropjle, through 

 which " it may be drawn by a slight traction completely out of the 

 nucleus, along with the suspended embryo." Both of these accounts 

 are quite at variance with the fact. The swelled knob, J, never ex- 

 tends beyond the micropyle, but is in every stage included within the 

 coats of the nucleus, and from this knob are sent forth two filamen- 

 tous processes, of which only the posterior one [c) has hitherto been 

 noticed. This process does not pass out at the micropyle, but perfo- 

 rates the coats of the ovule immediately below it on the outer side. 



Fig. 4. — Longitudinal section of the ripe seed : — d the micropyle; /the exti'einity of the anterior process 

 of the suspensor. The posterior process is continued from b to g ; e the embryo. 



The anterior process («) fii'st pushes its way through the substance of 

 the neck of the ovule ; then tlu'ough what may be considered the pla- 

 cental tissue. It then reaches, or forms for itself, in the substance of 

 the carpellum, at its inner angle, a canal which lies parallel with and 

 very near to the common axis of the three carpella ; and passing along 

 this canal the process arrives at the lowest point of attachment of the 

 carpellum with the receptacle where there exists a small pore. The 

 carpellary integument would thus seem to constitute an essential part 



4f 2 



