60 DR. J. D. HOOKER ON THE GENERA AND SPECIES 



earliest sta^ c at which I have examined this plant is that of an amorphous cellular hemi- 

 spherical mass nidulating within and almost enclosed by the bark of the roots, iato the 

 wood of which it had penetrated, its cellular tissue being in immediate contact with the 

 wood-fibres : in this early stage it contained no vascular tissue and no traces of lobing or 

 of peduncle. When fully developed, this species so entirely resembles K. Guyanensis in 

 all but the natm-e and position of the volva or involucre (here i-educed to an oblique ring 

 about half-way up the peduncle), as to preclude the necessity of a detailed description, 

 and I shall proceed at once to some remarks on its structtire and development. 



The structure of the wood of the rhizome is essentially the same as that of the preceding 

 species. The vascular bundles of the peduncle are, however, more regularly disposed. 

 The young peduncles arise as buds from circular depressions in the rhizome, and do not 

 exhibit any volva or involucre. 



The scales on the young capitulum are all ascending and imbricating, and do not 

 assume the peltate and hexagonal form till a subsequent period. 



At the base of the capitulum are seen several rows of small conical protuberances that 

 appear to be undeveloped bracts. The annular projection which represents an involucre 

 never to be developed, is situated close under the capitulum for some time after the 

 lengthening of the lower part of the peduncle ; finally, however, the upper part of the 

 peduncle elongates most, and the annulus hence occupies a middle position upon it. 



The articulated threads of the surface of the capitula are stout, of several collateral cells 

 in breadth, and the upper eight or ten cells are smaller and become sphacelate very soon. 

 In a young state these terminal cells are seen to contain several nuclei, and they are all 

 enclosed in a membrane of excessive delicacy, forming a balloon around them ; they appear 

 much whiter than the other cells at this early period, but afterwards are darker and 

 become minutely wrinkled on the surface. 



At the earliest period at which the female flowers can be recognized, they appear as 

 minute 2-3-lobed cellular papUlse, broader than long, upon the surface of the capitulum. 

 Wlien 2-lobed, the lobes (which lengthen into styles) are widely separated at their bases 

 and diverge. As they increase, the lobes lengthen and approximate ; and if a third one be 

 present, it is the middle one of the three that is suppressed. 



At this period the perianth is not distinguishable, but appears after the flowers have 

 assumed their complete form, as a broadly campanulate 2- (sometimes 3-4- ?) lobed 

 superior calyx, much larger in proportion to the ovary than at any future period ; the 

 styles at the same time have approximated, and appear united at their bases into a conical 

 body surmounting the ovary. The full-grown ovary differs in no important particular 

 from that of the preceding species. 



