20 
existence of beaks in the thecae of Andromeda Mariana. Fig. 7 
illustrates a stamen from a rather young bud. The beaks, here so 
distinct, become less so as the flower matures, by their close approxi- 
mation and adherence below. In the other species of that section, 
they are, even in the young state, beakless as described. 
We now come to consider a very different condition, best seen m 
Pyrola. Fig. 8 represents a stamen from a moderately developed bud 
of P. roiundifolia. Here we see the same folding of the filament (re- 
version) as in the first class of cases, the poriferous extremities of 
the thecse pointing downward, but at the same time inward. In 
the last particular they differ from all the above genera, and point to 
the inversion which is clearly shown in Fig. 9. In this case, one 
theca has been torn away, so "that the direction- of the fibres of the 
filament can be plainly seen at their actual point of insertion, again 
turning upward to become continuous with the connective. If, with 
a needle, we attempt to strip off the filament in a downward direction 
at the point a, we succeed in getting a transverse fracture at_ that 
point; while the sam© experiment with an anther of ^r^/z/z/i- similarly 
prepared results in the splitting off of the filament to tlie extreme end 
Fig. 10. 
of the connective. 
Fig. II 
Fig. 12 
Fig. 13. 
^ It is cleariy, then, this folding of the filament, 
hidden between the thecse, that produces the inversion proper, and 
proves the anther to be erect in the bud, and the pores to be basal; 
notwithstanding that in the mature flower they are uppermost. As 
to the direction of the pores, it is evident that, pointing inward m 
the bud, their normal position, they must point outward in the flower, 
when the inversion has become effective by the disappearance of the 
counterbalancing reversion. 
This extrorseness of the pores is permanent in Pytola, but not in 
Clethra or Chimiphila, and in that genus it is overcome, at least par- 
tially, by the unilateral depression of the anthers. But in CMhra 
(Fig. 10, stamen from bud, Fig. 11, from mature flower) the beaks 
are elongated, divergent and twisted, so that the pores face laterally 
ornearly so. In ClnmapJiila (Fig. 12, bud-stamen of C. njacidata, rig- 
13, flower-stamen of the same), the pores are practically introrse 
The manner in which they have become so is not clear, for in my 
only bud I found them extrorse in their reverted condition, and \ 
is not known to how early a period this relation could be traced. 
Some one will probably find a study of the buds of Rhododendron 
very interesting during the coming season. 
