THE CYMOSB TYPE. 157 



explained, except through full details : such as flowers standing 

 by the side of a leaf, or a small leaf by the side of a larger one, 



The transverse or oblique position of secondary axes or peduncles, as in 

 Eichler's first two species, brings the flowers of the false raceme or spike 

 out of line of the sympodial axis and bracts, neither in the axils, as in true 

 racemes, nor opposite them, as in the Rhipidium and Drepanium, but on one 

 side of this plane or the other. This is most common in Dicotyledons (in 

 Drosera, Sedum, Sempervivum, and Hyoscyamus, in Borraginaceae and 

 Hydrophyllaceae, &c.), and is not rare in Monocotyledons, especially with 

 tristichous phyllotaxy, as in Tradescantia. In the Bostryx, Fig. 299, the 

 bractlet is anterior or falls on the same side as the bract, or, in other terms, 

 the successive bracts are all on one side, the inner side, of the helix ; and the 

 Drepanium (Fig. 301) is like it: this is the helicoid cyme of Bravais, &c., 

 and its flowers are commonly one-ranked. In the Cincinnus or true scorpioid 

 cyme (Fig. 298), and equally in the Rhipidium (Fig. 300), the bracts fall 

 alternately on opposite or different sides of the sympodial rhachis, because 

 the single bract (u 1 ) of each successive secondary axis (a') stands next the 

 axis (a) and over against the bract (b) of the generation preceding. The 

 flowers in these generally fall into two parallel ranks (conspicuously so 

 when crowded) on the upper side of the rhachis, on which, in the cincinnus 

 or true scorpioid cluster, they are usuallj r sessile or nearly so (or spicate), 

 as is well seen in Heliotropes, and in very many Borraginaceous and Ilydro- 

 phyllaceous species, in Houseleek, Tradescantia erecta, &c. This comes 

 through antidromy , that is, the phyllotaxy of 

 each successive axis of the sympode (with its 

 one bract, or by suppression without it ) changes 

 direction, from right to left and from left to 

 right alternately. Fig. 301 is a plan of this 

 two-ranked unilateral arrangement. When not 

 too crowded, both Cincinnus and Rhipidium 

 are apt to have a zigzag rhachis. 



These two last-mentioned kinds are so gen- 

 erally alike in character, as are equally the 

 Bostryx and the Drepanium, that the four spe- 

 cies may as well be reduced to two. As these 

 severally include the scorpioid and the helicoid 

 uniparous cymes of modern anthotaxy, these 

 terms may be retained to designate them. Or, 

 if other terms in use be preferred to scorpioid 

 and helicoid, the form with two-ranked flowers 

 may be denominated Cincinnal, that with single- 

 ranked Bostryc/widal. But in neither type is the rhachis always coiled 

 up, although commonly more or less so in the undeveloped state. 



While these forms generally imitate racemes or spikes, it will be noted 

 that Fig. 300 specially imitates a corymb in form and in seeming acropetal 

 or centripetal evolution. And when, as in this figure, the bracts are all 

 absent, no obvious external difference remains. 



FIG. 305". Ground plan (from Eicbler's Bluthendiagramme, i. 38) of the scorpioid 

 inflorescence of Tradescantia erecta, between bract (B) below and axis (a) abo*ve: 

 I., II., III., &c., the successive flowers: v l is the bractlet of the first and bract of the 

 second flower, and so the others in succession up to v s and a small undeveloped one 

 beyond The figure 1 affixed to each flower indicates the first floral leaf. 



