THE ANDliCECTUM OF MENTZELIA. 211 
those of the third degree i and lastly, some five or six still lower ranges, 
in centripetal succession, each consisting of twenty stamens, or there- 
about, which alternate pretty regularly with the members of the range 
preceding them. It is to be noted, that in such an arrangement of parts, 
it is in the fourth range that there first appear stamens superposed to 
the petals, and exactly intermediate between any two stamens of the 
first degree. This will be sufficiently apparent by reference to the 
diagram wliich I have constructed, of the androeciura of Mentzelia^ 
and which represents the arrangement as satisfactorily as the difficulty 
of projecting, on a plane, points disposed on what approximately is 
the inner surface of a hollow cylinder, will allow. I have sometimes 
observed a slightly different arrangement, in which there ^vt fifteen 
instead of twenty stamens in the third range, a single stamen replacing 
the pair superposed to each petal, so that the first stamens which 
are exactly intermediate between any two primary stamens are thus 
in the third instead of the fourth range. I mention this latter ar- 
rangement thus particularly, because it is strikingly like that in Rosay 
to which I shall afterwards refer,* 
Almost immediately after the stamens begin to appear, the central 
portion or bottom of the receptacular cup becomes more markedly and 
abruptly depressed, forming a narrow funnel-shaped cavity, the rudi- 
ment of the inferior ovary, which is distinctly defined at its margin 
from the outer and upper portion of the receptacle upon which the 
stamens make their appearance. Around the margin of this central 
funnel-shaped depression three semilunar elevations are developed. 
These three processes are the carpels, which afterwards form the style. 
They are developed at a very early period, the time of their appearance 
being almost coincident with that of the ovarian cavity itself. I have 
seen them distinctly present when the stamens of the third degree had 
just appeared as very faintly marked mammilla?, so that it is almost 
certain that they are developed before these stamens. Probably they 
* Payer gives no figures of Mentzelia^ but he describes the appeai-auce of the 
stamens as follows :— * Cinq se montrent d'abord alternes avec les petales ; 
elles sont bientot suivies de dix autres placees deux par deux de chaque cote 
des premieres et un peu plus has, puis de qninze, puis de vingt-cinq, etc., de 
fa9on que rentonnoir floral en est promptement tapisse." As stated above, I 
have sometimes observed fifteen stamens in the thu-d range. I am unable, 
however, to say whether or not, in this arrangement, there are so many stamens 
as tweutj.five in any of the succeeduig ranges ; but I should be inclined to 
doubt that there are so many. 
