Griggs: Characters of the Platanaceae 391 



before flowering time its cells are dead, leaving it simply a hairy 

 fringe. Beyond these scales the staminate flowers are entirely 

 naked until long after the stamens are completely formed. When: 

 growth is resumed after the w^inter, however, the receptacle, which 

 has remained flat up to that time, begins to swell and forms bulg- 

 ing protuberances between the filaments (fig. 2, O), In cross-. 

 section (fig. 3) these are seen to be alternate with the stamens. 

 They are the structures which have been called vestigial carpels. 

 Sections of the carpellate head (figs. 5-7) show an exactly 

 similar condition. The shortened closely crowded peduncles are 

 surrounded by vestigial sheathing bracts. The folded carpels, 

 usually 6-9 in number, are located rather indiscriminately over 

 the upper surface of the receptacle but with their inner open sides 

 facing the center (fig. 6). Beside these the only other floral parts 

 present are three or four hairy staminodes (j*, figs. 5-7), which fold 

 over the carpels and protect them until just before anthesis. 

 While the homology of the outgrowths from the receptacle of the 

 staminate flower to carpels is somewhat uncertain because of their 

 rudimentary condition, there is little doubt that these appendages 

 of the carpellate flower are homologous with stamens. They 

 develop before the carpels, have a strong vascular supply, and are 



+ 



even sometimes quadrilateral. 



Such being the condition of the flowers, one turns with wonder 

 to the accounts of those authors who give not only descriptions 

 but figures of the petals and sepals. As has been said the aqgount 

 in the Pflanzenfamilien is the source of all the later descriptions. 

 But this is itself a compilation from the work of Schoenland (7), 

 from which its figures are copied. Schoenland describes the per- 

 ianth very circumstantially and in considerable detail. According 

 to him, I translate (p. 310): ** The flowers are perigynous ; they 

 possess always two regular, alternating, characteristically different 

 circles of involucral leaves each part of which may however show 

 considerable variation. I have no hesitation therefore in desig- 

 nating the outer as calyx and the inner as corolla. . . . These 

 circles are 3-6-parted (perhaps also to 8-parted). The numerical 

 relations may vary in one and the same inflorescence. In the 

 staminate flowers the 4-parted condition Is most frequent." He 

 figures a tetramerous pistillate flower with sepals, petals, stami- 



