430 TRANS. OP THE ACAD. OP SCIENCE. 
tal placentae and one-celled capsule) or on a fold which 
extends to the centre and forms the dissepiments (central 
placentae and three-celled capsule); shorter dissepiments 
make semi-trilocular capsules. Very fragile dissepiments, 
which break off when the capsule opens, leaving the placentae 
central and detached (septifragal dehiscence), are found in 
J. repens (therefore generically distinguished by Desvaux), 
and to some extent also in J. Parryi , J '. patens , and J. 
setaceus. The placentae of J. Boemerianus are enormously 
developed into a spongy mass, which fills the greater part 
of the capsular cavity. 
The capsule opens almost always from top to the middle or 
to the base; only in some of our species with subulate capsules 
{J. scirpoides, J. nodosus) the separation of the valves com¬ 
mences in the middle, while at the top they remain united 
for some time. 
Seeds. —The seeds, when perfectly ripe, furnish some of 
the most interesting and constant characters, but they are so 
small and their markings so delicate that only a strong glass, 
or, better, a microscope with a magnifying power of fifty or 
sixty diameters, will properly exhibit them. It may not be 
useless to caution botanists not to be deceived by seeds 
loosely lying about with the specimens, as they very often 
will be found mixed. 
The seeds are ascending or (the elongated ones) more or 
less erect, with a lower end at the insertion of the funiculus 
and an upper one at the chalaza, both ends united by the 
raphe and often by a distinct fold of the testa. The seeds 
are usually obovate or oblanceolate, thicker at the upper than 
at the lower end, mostly terete, or, in rare cases (J. trifidus ), 
angular, when a few large seeds are pressed upon one another. 
The ends are sometimes obtuse (J. bufonius), but commonly 
either abruptly or more gradually pointed, apiculate or even 
fusiform (J. pallescens , nodosus, scirpoides). V ery frequently 
the testa is slightly elongated beyond both ends of the body 
of the seed and forms a small, membranaceous appendage (P 
effusus, tenuis , marginatus) ; in such seeds the longitudinal 
fold of the testa, mentioned above, also becomes, more dis¬ 
tinct. In many, and apparently more in American and in 
alpine or arctic, species (P. JDrummondii , Greenii, Canaden¬ 
sis, etc.) these appendages become more conspicuous, and 
extend beyond the seed itself as empty, shrivelled, tail-like, 
white, or, rarely, reddish sacs. Such seeds have been called 
scobiform; their seed-coat is more loosely adhering and some¬ 
times (P. stygius) can be readily removed. This elongation 
of the testa is of great diagnostic value, but the absolute or 
proportionate length of the appendages is extremely variable; 
even in the same capsule I find the lower seeds with shorter 
tails than the upper ones, and in J. Canadensis we see forms 
i 
Botanical 
copyright reserved garden 
