ee ENNRENTEUmDeUM NF 
PREFACE. XXXVII 
nose the foramen is uppermost and the raphe conse- 
quently lowermost ; this was first pointed out to me 
by Mr. Brown, who immediately detected, that in my 
drawing of Humboldtia published in Dr. Wallich’s PI, 
Asiat. Rariores, I had reversed the situation. The same 
occurs in Asclepiadew and Scrophularinex, and proba- 
bly in most cases. In others again the foramen or raphal 
surface is lowermost, and in Henslovia it is indifferent/y 
lowermost or uppermost. 7 
In some instances in which the nucleus coheres/ in- 
timately with the outer integument, it seems run 
rather between the two than in the integument, this is 
very evident on making a transverse section/ of the 
ovula of some Composite. 
There is no appreciable difference of any eni 
between the development of the ovula in the/ two great 
. divisions of Phænogamæ, excepting so far as the em- 
bryo is concerned. In both, the first steps in forma- 
tion of this are the same. The same may be said of the 
Gymnospermous division of Dicotyledons, so far as I 
can judge from Cycadez and Gnetaces, the only two . 
orders Ihave had an opportunity of exaitsininz it in. 
I object to M. Mirbels' names which are founded ol _ 
Numerical considerations ; because in the first place, me 
the precise number of coats is not always developed, 
and one cannot well call a coat the quartine when it is 
only the third. The names founded either on the parts 
from which they originate, or especially on physiological 
points, carry a certain amount of information with 
them, hence for the tercine T propose, that of nucleary 
membrane, and for the quartine 1 adopt M. Brong 
