296 BOTANICAL GAZETTE [NOVEMBER 
at the base and extends toward the apex (fig. 1). The individual 
epigynous flowers appear as an undifferentiated mass of cells, 
somewhat rounded at first (fig. 2), but soon broaden out, so that 
the individual sepal primordia forming the calyx (which is well 
pronounced in Eryngium yuccifolium) are distinctly visible (figs. 
3, 4). This is soon followed by a perfect and regular acropetal 
succession, presenting the sequence sepals, petals, stamens, and 
carpels, in perfect accord with the account given by HANNAH (31) 
for Sanicula marilandica; by PAYER (56) for Heracleum Sphondylium 
(and other species of Heracleum), Carum, Aegopodium, Anethum, 
Phellandrium aquaticum, etc.; and by SIELER (62) for Heracleum, 
Sphondylium, Chaerophyllum bulbosum, Cicuta virosa, Daucus Carota, 
Peucedanum cervaria Lat., Angelica sylvestris, etc. (figs. 2-7). Itis 
well to note, however, that SEILER (62) is more concerned with the 
study of the sequence of the appearance of individual members 
within a cycle than with the sequence of the cycles themselves. 
JocHMANN (38) had noted that, although the calyx primordia 
make their regular appearance in Aegopodium, etc., they fail to 
continue in their development (“‘in pristino statu remanent, et quo, 
magis flos accrescit, eo magis evanescunt’’), and thus apparently 
simply remain as calyx teeth, and in many genera even these are 
obsolete and hardly distinguishable (figs. 9, 10). 
Ovules 
The carpels are distinctly two in number at first, but soon 
unite along their inner face, so that in cross-section they appear to 
be semicircular (fig. 7); or as PAYER (56), JOCHMANN (38), SEILER 
(62), and CAMMERLOHER (8) would have it in the forms studied 
by them, “semilunar” in shape. In each of the four free ends of 
the coalesced carpels an anatropous ovule begins to develop (figs. 
11, 12), one of which soon stops, however, while the other continues 
in its normal development. This results in a hanging anatropous 
ovule, with the raphe turned inward and the micropyle outward 
(fig. 14). This seems to be quite general for the entire family, for 
it has been found in all the species studied, and it is in accord with 
the account of CAMMERLOHER (8), who studied the “Samenanlagen” 
in thirty-seven genera and forty-five species. Moreover, JOCHMANN 
