8 
CONTRIBUTIONS TO BOTANY. 
longitudinal lines are seen upon the face opposite to the axis 
of the capitulum, which probably indicate the line of junction 
of two carpels, united thei’e by their margins, without any intro- 
flexure or tendency towards forming a dissepiment ; and it is 
probable that branches of the corda pistillaris from each stig- 
matic lobe run along these sutural edges of the carpels, as in 
the Capparidacece for instance. From the same circumstance 
we may also infer that the normal condition of the ovary is not 
2-locular with an intervening dissepiment ; for in such case the 
suppressed cell and the axis would be represented by a single 
longitudinal line. This inference is of course only hypothetical, 
but the suggestion is worthy of being kept in view. 
In Calyceracece the flowers in the same capitulum are not all 
fertile; for many of them are sterile and polliniferous, which are 
promiscuously mixed with the fertile or hermaphrodite ones. In 
Acicarpha, however, there is some exception to this rule ; for the 
superior or more central florets are all sterile, while the more 
external series are hermaphrodite and fertile. 
I have observed in Nastanthus, where the florets are promis- 
cuously intermixed, that the flowers first produced are not per- 
fect ; the tube of the corolla is considerably elongated into a 
very slender tube, on the outside of which are seen five promi- 
nences indicating the five transparent areolar glands, the tubillus 
within being very short, the filaments distinct, and the anther- 
cells, which are almost void of pollen, being almost, if not quite, 
free ; the segments of the border are of mueh thinner consist- 
ence, and of a much greener hue : in these cases the globose 
stigma is fully developed on the summit of the clavate extremity 
of the long style, and the achaenium yields a perfect seed. In 
the flowers last produced, and intermixed with the former, the 
tube of the corolla is thick, only half the length of the others, 
and so much thickened that the areolar glands become wholly 
immersed, and are not perceptible; and the segments of the 
border here exhibit the appearance of the gibbous double laminae 
before described; the anthers, almost obsoletely polliniferous, 
are nearly free ; the style is only slightly swollen at the apex, 
and deficient of the globose stigmatic expansion ; the achaenium, 
though attaining its full growth, does not always produce per- 
fect seed ; the corolla, in such instances, generally persists upon 
the achaenium. Other flowers, again, are produced in an inter- 
mediate state, the achsenium maturing its seed; but then the 
stigma is always fully developed, as well as the anthers, which 
are half united at their base into a syngenesious ring, and the 
corolla usually falls off soon after the period of impregnation. 
In Boopis, Gamocarpha, and Nastanthus, the calycine lobes 
are deeply concave or semi-narfcular, owing to their involution 
