212 BOTANICAL GAZETTE [SEPTEMBER 
tained two stamens, opposite the sepals (here doubtless the dorso- 
ventral pair), and in the middle a minute and obscure rudiment of 
still another flower (/). In the whole axis we have thus a case of triple 
prolification, together with distinct ecblastesis from each sepal of the 
middle pair. 
The upper flowers of the raceme, which we have been following, 
although modified in various degrees, showed a greater and greater 
simplicity of structure, as will be seen from the three represented in 
figs. 7,8, and 9. The last shows the simplest form of flower observed. 
Its envelope consisted of two sepals with which alternated two obscure 
microscopic rudiments, doubtless the inner sepals. The andrcecium was 
formed of two stamens opposite the developed sepals, and the gyn- 
cium of two carpels of normal position. ig. zo shows a diagram of 
the flower. For the sake of comparison the normal flower from one of 
the terminal racemes of the same plant is shown in fg. zz. In this 
the minute rudiments of petals may be seen alternating with the sepals, 
and one of the two stamens is just visible protruding from behind the 
ventral sepal. The sessile normal capsule of orbicular contour of 
course presents a striking contrast to the deformed stipitate gyncecium 
of the modified flowers. The specimen shows an apparently sound 
and normal root and in general shows no cause for deformities 
described. The plant is preserved in the Herbarium of the Middlesex 
Institute at Malden, Mass.—B. L. Ropinson, Harvard University. 
