455 
in the Stock (.Mat thiol a incanci). 
tissue, and forming the farther central end, are larger, colourless, elongated, 
irregularly shaped cells, which later also become thick-walled and lignified. 
The two ingrowing septa eventually meet, the lining epidermis disappears 
from the two contact surfaces which unite, and the partition is complete. 
Later the cells at the point of junction become converted into a longitudinal 
strand of fibres, a character of importance from the systematic point of view. 
3. Interpretation of the siliqua construction . 
The commonly accepted view of the Crucifer siliqua, as mentioned 
above, is that it is composed of two lateral carpels forming the two valves, 
and that the two smaller arcs, together with their inward prolongations, 
represent the incurved fertile margins of these carpels, which finally become 
united centrally by a later developed tissue, the origin and significance of 
which, on this view, remain unexplained. There are also admittedly other 
features which on this interpretation present a certain difficulty, the chief 
being firstly, that when the valves break away the small arcs, i.e. the 
supposed fused edges (placentae), are left behind on the partition ; and 
secondly, that the position of the two stigma lobes, when these are distinct, 
is in nearly all cases over the sutures. Thirdly, there is the fact, which is 
clearly shown in the drawings of Payer, Eichler, and others who have 
investigated the early development of the gynoecium, that the placental 
commissures from the first moment of origin are larger than the carpels of 
which they are supposed to be merely the margins. In order to meet the 
stigma-position difficulty it was suggested by Lindley 1 (1828) and also by 
Kunth 2 (1831-3) that the supposed conjoined edges or placentae really 
represent independent carpels, the siliqua thus being composed of four 
members, a view adopted also by Godron. 3 Eichler, 4 though admitting 
that certain appearances, especially in the Papaveraceae—another family 
which presents the same problem—lend support to Lindley’s idea, never¬ 
theless regards it as preferable to accept R. Brown’s 5 conception of the 
commissural stigma, and to consider the number of carpels in the Cruciferae 
as two. Benecke, 0 Chodat, 7 and Celakovsky, s surveying the arguments and 
evidence afresh, have severally endorsed Eichler’s formula, either expressing 
the opinion that the construction of the Cruciferous flower is dimerous 
throughout (Benecke and Celakovski) or that the full gynoecium ground- 
1 Bot. Register, vol. xiv (= vol. i of 2nd series), text accompanying PI. 116S ( Eschscholzia 
californica). 
2 Handbuch der Botanik : also fiber die Bliithen- und Fruchtbildung der Cruciferen. 
3 Ann. Sci. Nat., Bot. v, 2 e ser., p. 293. 4 Bliithendiagramme, ii, p. 192. 
5 PI. Jav., note, p. 108. 
6 Zur Kenntniss des Diagramms der Papaveraceae und Rhoeadinae (Engler’s Bot. Jahrb., ii, 
1882). 
7 Neue Beitrage zum Diagramm der Cruciferenbliithe (Flora, lxxi, 1888). 
8 Das Reductionsgesetz der Bliithen (Sitzb. d. K. Bohm. Ges. Wiss., No. 3, 1894). 
