62 
Indeed, Wood in his Class Book, placing the plant in the genus 
Paronychia, calls it Canadensis, perhaps compelled to this by the 
fact that Paronychia already held a species dichotoma, quite another 
aftair. But it the two species be distinct, as I am inclined to be- 
lieve, it may be a matter of doubt as to which form is entitled to 
take the Linnean name. But as Nuttall first discriminated between 
the two, it would be just to retain his name A. capillacea for the 
one, and to give the name dichotoma to the other, thus avoiding 
the doubt. Joun H. RepFiep. 
Philadelphia, Oct. 22. 
§ 62. Pontederia cordata, L-—A season or two back a clump of 
Pickerel-weeds engaged the attention of Mr. Hall and myself by an 
appearance of di- or trimorphism in the flowers, and we gathered a 
number of heads for examination at more leisure than we then had. 
It was, however, only late this autumn, too late to get fresh speci- 
mens, that I renewed my examination, and in the dried flowers I 
did not well make out what it was that struck us in the fresh ones, 
and unfortunately Mr. Hall is not here to give me the benefit of his 
recollection. H. Muller (Befrucht. der Blum. p. 62.) gives Kuhn ( Boé. 
Z., 1867, 8. 67) as his authority for stating that Monochora, L., has 
cleistogamic flowers ; and states that his brother has found in South 
Brazil two species of Pontederia probably trimorphic. I hope to 
resume the subject next season, but in the meantime wish to call 
attention to the result of the examination of the dried spikes, as it 
revealed some points of interest. 
The spike of Pontederia is compound, the spikelets being arranged 
in the 4 system common toendogens. The main spike begins to flower 
at the base, but, as the flowers of the spikelet develop successively, 
mature and immature flowers may be found all along the spike, when 
in full bloom, giving it its somewhat ragged appearance. There 
are three (or perhaps sometimes four) flowers in each spikelet, but 
to what extent they all develop I am in doubt. 
In all the flowers which had matured and coiled up, I found the 
style as long at least as the longer stamens, but. in all the other 
flowers, whether fully opened or apparently about opening, the style 
was intermediate in length between the two sets of stamens. This 
lengthening of the style as the flower matures occurs in other 
plants (for example Epiphegus, Menyanthes,) and in some cases 
might be explained by its partaking of the general growth of the 
pistil, but in this case and that of the sterile flowers of Epiphegus, 
this solution is not so satisfactory. 
Of the six stamens, three on long and three on short filaments, I 
uniformly found the anther cells empty or nearly so in the tully 
opened flowers, but in those apparently nearly ready to expand the 
anthers seemed just mature and shedding their pollen, all the six 
equally. In the unopened flowers all the stamens held their heads 
erect, but in the opened flowers the shorter three uniformly had 
their anthers turned down. These unopened flowers seemed gen- 
erally, if not always, the second one of the set on the spikelet, but I 
have not examined this sufficiently. 
The most remarkable point noticed was the difference in the size 
