306 



lamp 1 have seen, especially as the large yellow flower may be consi- 

 dered lo represent the flame. The rays are half an inch long, clavate, 

 not hollow, but composed of about six series of large diaphanous cel- 

 lules. The cellules are convex on the surface, giving the rays a 

 papillose appearance, hexagonal, pale green, with pink insterstices. 

 The ra^^s are trifid at the extremity; segments short, twice dichoto- 

 mous, the last divisions capillary, rarely sacciferous." 



" I lately found on the leaves of No. 960, what I at first took for a 

 new Rafflesiacea, and the resemblance to some Apodantbus is indeed 

 most striking : it was only by careful examination that I satisfied my- 

 self it was really produced by an insect. The perianth (for such it 

 seems) is green in the earliest stage, changing to pink, and afterwards 

 to dull purple, tubular from an oval base, from one to two lines long, 

 and the tube a third of a line broad, hairy within and without with 

 spreading white hairs (though the leaves are nearly smooth) ; the 

 mouth expanded, 2 — 5-lobed, sometimes dimidiate ; ovary inferior, 

 1-celled, with one or two pendulous ovules. But these ovules are the 

 true eggs of an insect, for, by examining individuals in progressive 

 states of development, I have traced the formation from the egg of, 

 first, a minute fusiform annulate body, and, ultimately, of a perfect 

 insect with legs and wings. To make the resemblance to a flower 

 more striking, there appears, beneath what I have called the perianth, 

 what seems to be a calyx of four or five erect triangular brownish 

 sepals ; but these are really only the torn cuticle by the protrusion of 

 the perianth. 



" To explain the form assumed by these excrescences, may we not 

 suppose there has been an attempt to reproduce the tubular 5-lobed 

 calyx of the species (which belongs either to Inga, or to some allied 

 genus) .? The juices of a plant, when diverted from their ordinary 

 channels, must still go on forming tissue according to some law origi- 

 nally impressed on the species; and I have seen modes of develop- 

 ment follow the puncture of an insect, such as in general only long 

 cultivation calls forth. On the same leaves were the nidi of another 

 insect. These were scarcely a line long, globoso-urceolate, regularly 

 20-striate, containing eggs in the concavity as in the other. They 

 might easily be mistaken for some epiphyllous fungus. I enclose 

 specimens of these productions, and I will afterwards send you a 

 larger species." 



