IN GNETUM, 167 
2, Long section of same; a. long funicle; a, raphe; 5. 
testa; c. tegmen ; d, nucleus. 
3. Portion of placenta, with ovula, at about similar pe- 
riod as No. 10 of Fig. 3; a. placente; 5. their ap- 
proximated margins; c. c. c. hila; d. d. d. forami- 
na somewhat renioved from their original positions ; 
into them, a tube is seen to pass from the placenta. 
Gnetum. The sudden appearance of the third envelope, 
whatever its origin may be, is very remarkable. Thus at the 
time at which impregnation may be supposed to take. place, 
that is during the dehiscence of the anthers, there is no trace 
whatever of it. After the fall of the male flowers, and when 
the ovuleis as large again, we find it perfect, with its styliform 
prolongation just exserted. I believe it to be either a new 
membrane or a formation from the pre-existing secundine. 
It is distinct from the nucleus as far as its origin is concern- 
ed, for this last alters but little, and is always recognisable, 
changing but very slightly in form, the quintine, now quar- 
tine, being visible too. i 
The fact of this apex of the secundine projecting beyond 
the exostome during the dehiscence of the anthers, is another 
remarkable cireumstance, especially when coupled with the 
future and much greater exsertion of the new membrane. 
The study of the impregnation will be highly interesting, 
and will for obvious reasons, as well as that of Coniferz, 
throw vast light on this intricate phenomenon. 
How does the pollen exert its influence which is general 
enough, since many ovules in a spike are always impregnated ? 
The great exsertion of the apex at the third membrane, 
and its sphacelation would seem to indicate some co-opera- 
tion, but this takes place after the fall of the male flowers of 
the respective spikes. - Besides which, the quintine has al- 
ready appeared. : 
The whole organization is precisely the same in the three 
species. It is a curious fact that very little change in the 
position takes place in the male flowers, except n. those of 
G. rufipilosum, in which they become per anthesin exsent- 
ed as it were, fi 
