CICHORIACEAE. 61 
the segments were deep and narrow, as now seen in many a cam- 
panulaceous flower. 
The theory of the origin of a ligule from a regular pentam- 
erous deeply cleft or parted sympetalous corolla naturally pre- 
supposes one of two distinct modes of transformation. It is 
conceivable that, by the gradual congestion into a dense head of a 
loose campanulaceous inflorescence, the deeply cleft corolla might 
loose, one after another or little by little, all of its segments 
Save one; but we should not expect a ligule thus derived to be 
toothed at all at apex. And what is more, in the curious cam- 
panulaceoug genus Jasione, in which the flowers are crowded 
into a dense head, involucrate like that of composites, there is 
still no reduction or alteration of the corolla, this consisting of 
five narrow-linear equal segments; and the supposition that the 
ligule of the chicory came about by elimination of segments 
seems precluded. 
The other natural hypothesis, to me seems this : hae there was 
a cohesion of the five segments beginning at the tip and pro- 
ceeding downwards, until, by a natural tension, a rupture of the 
incipient apical tube by the uppermost of the five sutures fav- 
ored a complete union of the segments downwards by the other 
sutures, until finally the five-toothed—equally five-toothed— 
ligule become established. 
The anthological phases, of another campanulaceous genus, 
Phyteuma, illustrate remarkably well the possibilities of the 
descent of the chicory ligule along such a line, and by such 
gradual modifications of a regular aud deeply five-parted corolla, 
In some species of this genus there are the five linear rotate- 
spreading segments of a kind not unknown in other related 
genera. In some such it has long been noted by botanists that 
before the full expansion of the five narrow segments, they 
cohere lightly at tip, the expansion thus seeming to proceed from 
the base in such wise that at the tips the segments are finally 
forced apart by the tension to which they yield at last somewhat 
suddenly. In others of this genus, the tips are never sundered 
at all, but form a permanent five-toothed tube which reaches 
down to the middle of the corolla, or near it, or below it, and 
below this tubular part, the portion of the segments still 
