44 CALIFORNIA ACADEMY OF SCIENCES. [Proc. 3D Ser. 



that the so-called flower is really a branch system, arising 

 by a true dichotomy of the primordium in precisely the same 

 way that the latter arose from the dichotomy of the apex 

 of the main shoot of the plant. Although the matter was 

 carefully investigated there was no evidence that the flowers 

 ever arose as lateral branches of a common axis. 



The early stages of the development of the flower have 

 been investigated by Schumann (1892) and Magnus (1894), 

 but owing to the impossibilit}'- of tracing out the youngest 

 of the stages except by the aid of microtome-sections, they 

 failed to make out correctly the origin of the carpels, and 

 especially the real nature of the ovule. Both describe the 

 carpel as arising in the form of a ring-shaped wall, upon 

 whose inner surface the ovule is formed as a secondary 

 growth. A complete study of microtome-sections, through 

 all the younger stages of the flower, has convinced me, 

 however, that the ovule is really an axial structure, being 

 the termination of a branch, and the carpellar}- leaf merely 

 subtends it. 



If an exact median section through the youngest stage in 

 which the flower can be recognized is examined, it will be 

 seen to consist of a very slightly elevated protuberance (fig. 

 93, £>), which is the apex of the very short branch resulting 

 from the last dichotomy of the floral axis. On the outer side 

 of this is a crescent-shaped ridge, partially surrounding the 

 apex, in section appearing as a pointed, almost horizontal 

 protuberance (fig. 93, c). This is the carpellary leaf, but 

 as its position shows, it is a structure entirely independent of 

 the apex (o), or perhaps it may be better considered to be 

 an appendage of it. The true relation of the two structures 

 is still more evident in a later stage (fig. 94), where the 

 tissues of the apex (o) show all the characters of the ordinary 

 stem-apex. The base of the carpellary leaf extends around 

 the base of the shoot and is confluent to some extent with 

 it; but in no case do the two form a circular ridge of uni- 

 form height and thickness such as Schumann (1892, PI. V, 

 fig. 3) figures and describes as a circular carpel upon whose 

 inner face the ovule arises later. If we compare with the 



