GENUS MVS'rROPET.lLON IIARV. ( UALA Nol' I loKAClCK) . 2/9 



after its discoverer. As will be seen further on, I do not intend to 

 do that. 



Harvey-Ciibson arranges the floral characters of the three 

 species in a tabular statement which I venture to reproduce. 

 Before doing- so, however, I should like to refer to some of the 

 points of comparison in iJarvey-Gibson's table, as this will sim- 

 plify matters and save repetitions later on. 



Point 3. The Anterior Lobe of the Perianth of the Male 

 Flower. — In young buds of all specimens examined I found the 

 three lobes nearly of equal length and the perianth tube straight 

 (at that stage the three segments form a tube). When the time 

 of opening api)roaches, the claws of the two posterior lobes 

 lengthen more rapidly tlian that of the anterior one, thus pro- 

 duc'ng the downward curving of the flower. The tension between 

 the two longer and the one shorter segment becomes hnally so 

 great that the segments separate ; the blades of the two upper 

 segments generally remain comiected, while the free lower seg- 

 ment curves back. Naturally this lobe is shorter than the two 

 others, but the difference in length varies in the same spike. No 

 specific distinction can be based on this feature of the i)lant. 



Point 4. Lamina of tlie Perianth Lobes of the Male Flower. 

 — A distinction is drawn by Harvey between " slightly concave " 

 (M. Thoinii) and "very concave'" (M. Polemanni) . In all the 

 specimens examined, including Bolus, No. 7465, which has been 

 named M. PolenuDiiii at Kew, I found the lamina " slightly con- 

 cave " or just " concave," and Harvey's specimen, on which it 

 was " very concave," was just a slightly different form. Such 

 differences cannot have specific value when other characters vary 

 much more. 



Point 5. The Pollen. — Harvey describes the pollen grains of 

 both species as possessing " fluted angles," and figures them thus 

 for M. Polemanni. On none of my specimens (I may have 

 examined at least 50 from different localities, all in a fresh condi- 

 tion), were the angles fluted. Judging, however, from the outline 

 of the grains as figured by Harvey, it appears that these had 

 undergone considerable shrivelling and contraction (perhaps by 

 desiccation or by the action of strong alcohol), and that the edges 

 had become furrowed during this process. 



By staining the fresh grains I have found that the rounded 

 edges of the cubical grains (this is the most common shape) 

 contain two parallel strands, and in some shrivelled grains I have 

 observed a slight depression along the edges. Under the circvun- 

 stances one feels compelled to discard Harvey's figiu-e and to 

 look upon Harvey-Gibson's Fig. 6a, and Marloth, Tab. 40, Fig. 6, 

 as more correct. 



Point 10. The Shape of the Oz'ary. — This is said to be either 

 ovoid or oblong. Both shapes occur fairly often on the same 

 plant. 



Point II. The Stigma is described as "discoid," or "not 

 swollen," and for M. Sollyi as " trilobed." All the plants which I 

 have examined when in anthesis (before any male flower opens) 



