ON THE STAMINAL ARRANGEMENTS IN SOME SPECIES 
OF POTENTILLA AND IN NUTTALLIA CERASIFORMIS. 
By ALEXANDER DicksoN, M.D. EDIN. 
(Puare LII.) 
On examining, about a year ago, the flowers of Potentilla fruticosa, 
I was much struck with the disposition of the stamens. These are 
arranged in strongly-curved lines or festoons, each containing 4 or 5 
stamens, and extending from petal to petal. The convexity of each 
festoon is towards the centre of the flower, and there are no stamens 
superposed to the petals. I have since then examined the develop- 
ment of this andreecium, and, as might have been anticipated from the 
analogy of the rosaceous developments already observed, I find that 
each festoon the two stamens next the adjacent petals are the first ie 
veloped ; the. two or three forming the middle or lower part of the fes- 
toon appearing subsequently. It is very difficult exactly to observe 
whether or not the central stamen of the festoon, when this consists of 
5 stamens, is actually younger than those on either side of it. I have 
not been able with certainty to detect any decided difference of size be- - 
tween them; and the absence of the middle stamen at a given time 
does not afford any sure proof of its being a later development, as it 
not unfrequently never appears. udging, however, from the analogy 
of the other Rosacee, it may be considered almost certain that the 
central stamen of the festoon is the youngest. When the stamens , 
have all appeared, they, together with the “ petals,” form a pentagon 
of mammille surrounding the hemispherical termination of the floral 
axis. The petaline mammille form the angles of the pentagon, and 
are the oldest and largest; next in size and age are the stamens 
nearest the petals; and youngest and smallest are the two or three 
stamens in the middle of the sides of the pentagon (Plate LH. 
Fig. 5). Icannot but think that such an arrangement strongly con- 
firms the doctrine of rosaceous andreecia propounded in my paper on 
Mentzelia, etc. (Journ. of Bot. iii. p. 209) ; as I am unable to conceive 
of any possible explanation of such a festooned arrangement of stamens, 
unless we view the andreecium here as consisting of five compound 
and confluent stamens, the terminal lobe of each such stamen being 
VOL, IV, [SEPTEMBER 1, 1866.] T 
