154 REV. PROF. G. HENSLOW ON THE 
character of many, otherwise slender stalks, as of Tria, Pinks, &e., 
is attributable to the presence of this structure. In some cases 
collenchyma answers the purpose, as in Cyclamen; aud the 
curious way in which the peduncles of this flower collapse when 
fertilization has not been effected is a consequence of the with- 
holding of fluid; whereas in Ixia, if the flower be dead, the 
sclerenchymatous cylinder of the peduncle is sufficient to enable 
it to remain nearly as rigid as before. 
Another distinctive feature between the two classes resides in the 
way the pedicels are formed in an umbel from a common peduncle, 
This is well seen by comparing Ribes (xxv. 1-3; Pl. XXVIII.) or 
Primula (x1ix. 1-5; Pl. XXXI.) with Narcissus Tazetta (LVII. 
1-7; P.XXXIL). Inthe former, a portion of the fibro-vascular 
cylinder of the peduncle bulges out like a loop or horseshoe, as 
seen in a transverse section. The outermost cord or cords sepa- 
rate off and supply the bract, leaving two crescent-like portions, 
their concavities facing each other. These close up and form a 
small cylinder for the future pedicel which will issue from the 
axil of the bract. In some cases, if the peduncle contain a few 
isolated cords, and not a completely connected cylinder, the 
tracheæ of each of the isolated cords multiply by chorisis, spread 
out like a fan, until, by repeated division, the outer ones meet, 
and then collectively form a cylinder of themselves. This now 
supplies the pedicel. Other cords pass out and supply the bracts. 
Such a method obtains in Erodium cicutarium (xvi. 1-3; PI. 
XXVL); and finds a parallel, in the way the stamens are supplied 
with a central cord, in Euphorbia (111. 4-5; Pl. XXXI). 
In Endogens the peduncle, e. g, of Narcissus Tazetta (tvi. 1-7; 
Pl. XXXII.), contains a circle of large cords embedded in loose 
parenchyma which has generally a lacunain the centre. There is 
usually a circumferential series of smaller cords irregularly dis- 
posed. On reaching the solid node, the cords all follow oblique 
directions, and branch in an extraordinarily confused manner. 
They then gradually rearrange themselves into groups of variable 
numbers, such as five or six, or even so few as three if the pedicels 
are of small diameter. The fundamental tissue being at first 
uniform throughout, gradually becomes marked out into islets, 
which separate themselves by developing epidermides. Thence, 
the petioles, often angular in section from their origin, are 
perfectly formed. 
