364 THE AMERICAN NATURALIST. [VOL. XXXII. 
writers, Eichler has, after an admirable summary of previous work 
upon the subject, stated the simplest and perhaps most convincing 
plan. His diagram, which has in recent years met with pretty wide 
acceptance, is as follows: the calyx consists of two dimerous alternat- 
ing whorls ; the corolla of a single tetramerous whorl, of which the 
parts alternate with the sepals taken together; the andrcecium of 
two dimerous whorls (the members of the inner being doubled by 
division), and the gyncecium of two laterally placed carpels. 
This theory is too well known and has been too carefully grounded 
to need any explanation or defense here. In recent years, however, 
three more or less divergent views have been expressed by Klein, 
Celakowski, and Lignier. Passing over some slight points, one may 
say that the plan of Klein differs from that of Eichler in maintaining 
a tetramerous inner whorl of stamens and a four-carpelled gyneecium, 
in which not only the two valves, but also the two placenta-bearing 
columns of the replum, represent carpels. Celakowski, however, 
believes the andrcecium to be derived from two tetramerous alternat- 
ing whorls, the outer of which has lost two of its members by abortion. 
He agrees with Eichler and most of the earlier writers in regarding 
the gyncecium as fundamentally bicarpellary. 
Lignier’ has suggested a theory of which the ingenuity is only 
exceeded by the disregard for facts. He supposes the flower to 
consist of only four alternating dimerous-whorls. The first consists 
of the two outer sepals, which he believes lateral.. Then follow the 
two median sepals, which he regards as three-parted, the green sepal 
being the central part of each and the two adjacent petals being the 
lateral parts or lobes of the sepals. Similarly, the short stamens are 
regarded merely as the middle lobes of trifid members, of which the 
adjacent longer stamens represent the lateral parts. Even in the 
gynoecium Lignier endeavors to show connate three-parted members, 
since he regards the placenta as the central lobes and the valves as 
composed of the connate lateral lobes of two carpels ! 
The latest publication upon the cruciferous flower is that of 
Chodat and Lendner.? 
These authors have made a detailed examination of the floral 
development, especially as to the course of the fibro-vascular bundles, 
and devote some space to a refutation of Lignier’s theory, — a 
matter of no great difficulty for any one reasonably conversant with 
the early stages of the cruciferous flower. The argument is chiefly 
1 Compt. rendus. Acad. Sci., pp. 67 5-678, 1895. 
2 Bull. de V Herb. Boiss., V., pp. 925-938, November, 1897. 
