ME. E. TBIMEN ON THE STErCTURE OF BONATEA SPECIOSA. 157 



observe tliat botanists agree in grouping it witli Ilalenana, 

 Lindley remarking (' Gen. et Sp. Orchid.'), " Bonatea est Jlabe- 

 naria stigmate evoluti:ssimo." My object here is limited to the 

 illustration of the manner in which the structure of B. speciosa 

 ensures the removal of the pollinia by insects resorting to the 

 flower, and how fertilization ensues through the same agency. 



The general aspect of the complete flower may be seen in figs. 

 A. & B, of which the one gives a side, and the other a front view. 

 The sepals, which are pale green, present nothing remarkable in 

 form ; but the two low^er ones are sharply pointed and subfalcate, 

 vs-ith their edges curled inwards underneath. The petals, on the 

 contrary, are singularly shaped and greatly elongated. Each 

 upper petal is divided almost to the base, with the green posterior 

 portion standing erect, just within the margin of the upper sepal, 

 and meeting at its apex the same portion of its fellow petal; 

 while the white anterior portion projects almost at a right angle 

 with the posterior, cohering, for a short distance, by its inner 

 edge with the lower petal or labellum, and thence, w^idely di- 

 verging, becomes gradually broader, but somewhat abruptly nar- 

 rows at the apex, which is prolonged and strongly recurved. The 

 bright green labellum has a filiform dependent nectary as long as 

 the ovarium, and is tripartite, its central lobe being much angu- 

 lated, and depending between the lateral limbs, which project 

 outward, more after the manner of the anterior portions of the 

 upper petals. 



The very remarkable cohesion of so many organs wuth the la- 

 bellum has been correctly indicated by Mr. Darwin {op. cit, p. 304, 

 note) ; and a reference to Fig. D will further illustrate this point; 

 The long stigmatic processes (of which more hereafter) cohere on 

 either side, for rather more than half their length, with the uj>per 

 surface of the labellum; with its outer edges cohere, as haa 

 already been mentioned, the anterior portions of the upper petals 

 for barely one-fourth of their length ; and, finally, the inner edges 

 of the lower sepals cohere with its under surface for about oner 

 third of their length. It is thus clear that, as Mr. Darwin has 

 pointed out, a section through the base of the labellum would di- 

 vide these coherent parts also, passiag in ajjf^through no less than 

 ^even (or, regarding the lateral limbs of^the labellum as petaloid 

 anthers of the outer whorl, nine) disd/fct organs 



The rostellum forms a very con^Jicuous green hood or helmet, 

 constricted in its neck, and with the sides produced anteriorly 

 into two long arms or horns projecting in an upward direction 



