oe 
On a Dimerous Form of Pansy. 633 
Stamens four, in two lateral pairs, the two stamens of each 
pair being completely connate, their connectives having grown 
together, and the two pairs being slightly united by interlacing 
fringes of hairs, like the slight connexion between all the 
stamens of the ordinary flowers of the same plant. Anther cells 
introrse, like those of ordinary flowers. The pollen was mostly 
shed when the blossom was examined, and therefore could not be 
sufficiently scrutinized. What was collected appeared more 
prevailingly round than is usual in pansies. The filaments of 
all four stamens were spurred, like the filaments of the two lower 
stamens of ordinary flowers. Two of these spurs, 7.¢., one from 
a stamen of each lateral pair, passed into the spur of the lower 
petal; the other two spurs passing into the spur of the upper 
petal. The stamens had orange tips, and except as noted above 
resembled those of ordinary blossoms. 
Style as in ordinary blossoms. Stigma cupshaped, but with a 
straight opening, not one leaning downwards as in ordinary 
blossoms—at least this was the appearance when the flower 
was examined, but the edges of the cup were then slightly 
withered. Ovary a one-celled capsule of two carpels, many- 
ovuled, placentas parietal, and lateral, i.c., opposite the sepals. 
Dorsal ridges above and below, i.¢., opposite the petals. The 
carpels resembled those of ordinary blossoms, but there were only 
two instead of three of them. Ovules anatropous. 
Thus all parts of the blossom were dimerous, and the blossom 
seemed to approximate to the Fumariaceze, Cruciferee, and Cap- 
paridacew, except as regards the position of the carpels. If this 
abnormal blossom in one of the Violacese may be regarded as an 
instance of reversion towards an ancestral type, it would seem to 
indicate that the order of descent has been from a plant, some- 
what resembling our Fumariacez to our present Violacez, and 
not in the opposite direction. 
