142 BOTANICAL GAZETTE [FEBRUARY 
usual way; the floral bracts all develop into spines except those of 
the two inclosed flowers, which remain rudimentary; and the 
“‘Anlagen” of the two flowers which appear follow the normal 
sequence of development. The only respects in which this head 
differs from those of the typical Compositae are seen in the two beaks 
and the depressions which they subtend. The beaks cannot be 
considered as floral bracts of the subtended flowers, because the 
rudiments of these parts are present in the pits. Neither is it likely 
that they are the floral bracts of aborted flowers, for they are quite 
unlike the spines in form, structure, and development. The fact 
that they bear the spinelike floral bracts on their outer surfaces 
precludes their being interpreted as modified involucral bracts. 
Moreover, two whorls of involucral bracts have never been estab- 
lished for any of the Ambrosiaceae. Though some writers (13) 
have interpreted Jva as having a double whorl of involucral bracts, ° 
my study (7) has shown that the inner of these whorls might 
properly be regarded as the floral bracts of the pistillate flowers. 
It seems most reasonable, therefore, to interpret the beaks of 
Xanthium, not as modified bracts nor even as newly developed 
structures, but as portions of the receptacle formed by its upward 
growth, and slightly altered by proximity to the depressions. 
Under this interpretation the only modification required to trans- 
form a typical head of Compositae into a Xanthium-like bur is the 
formation of two depressions in the apical region. In this species 
it seems that these depressions arose, not by a sinking of the flowers 
into the receptacle as HoFMANN (6) suggests, nor by an intercalary 
growth in the surrounding regions, but rather by an arrest of 
development in the apex due possibly to contact with the tips of 
the involucral bracts at an early stage. 
In a recent treatise on sexual differentiation in plants (8) 
GOEBEL ascribes to modifications of nutrition, not only the origin 
and evolution of sex, but all the phenomena of hermaphroditism, 
dicliny, and dioecism. He surveys the entire plant kingdom and 
applies his interpretation to both cryptogams and phanerogams, 
and to gametophyte and to sporophyte alike. Numerous experi- 
ments and observations indicate that in many cases nutrition does 
determine the sex of an individual or part. But, especially in the 
