140 ME. C. B. CLA.B.KE OK INDIAN SPECIES OF CTPEEUS. 



Sect. L. Glomerati. 



Majusculi, foliati. Radix fibrosa, valida, rarius lignosa. Cul- 

 mus ssepissime solitarius, trigonus, apice triqueter. Umbella com- 

 posita aut decomposita, spicis numerosis, approximates, multi- 

 spiculosis. Spiculse multiflorse, compressas ; rhachillse alse con- 

 spicuse, persistentes. Glumae laxiuscule imbricatae, ovatae aut 

 ellipsoideas, obtusae. Stamina 3 ; antherae lineari-oblongae, muticae, 

 rarius apice obsolete cristatae, sanguineae. Nux cum |— | parte 

 glumae aequilonga, ellipsoidea aut oblonga, apice saepe angustata, 

 trigona, baud raro inaequalis, interdum curvata. 



From tins point till we come to tbe Exaltati there is a long 

 series of species wliich I cannot divide into any satisfactory 

 sections ; but, to avoid one inconveniently long string of species, 

 I have arranged them pretty much as Boeckeler has left them. 



Boeckeler has placed G. distans as an anomalous species, but I 

 cannot sort it satisfactorily even as a species from G. nutans ; and 

 Boeckeler calls Sieber n. 18, G. distans, whereas I think if there 

 is a typical G. nutans, that is it. Boeckeler places G. spectabilis 

 very near G eleusinoides, but arranges G. glomeratus itself with 

 0. Sorostachys (leucocephalus, Retz.), whicb I think is certainly 

 not the true affinity ; indeed G. spectabilis has been issued from 

 Kew as a mere form of G. glomeratus, and I think it is excessively 

 near it. 



Benth. et Hook. f. Gen. PI. iii. p. 1045, have followed Kunth 

 in making their large sections (subgenera) depend upon the 

 development and persistence of the wing of the rhachilla : this, as 

 stated in the introduction, I think is impossible to work ; and, so 

 far as it can be worked, leads to most unnatural combinations. 

 In species like 0. rotundus, of which we have abundant material, 

 it is seen that the wing is generally finally soluble : the point of 

 age at which it is soluble appears to be very variable, and in her- 

 barium specimens depends largely on the method by which the 

 specimen was dried. As regards the breadth of the wing, I doubt 

 its value as a subgeneric character : G. cephalanthus is arranged 

 next G spectabilis by Hooker f., and I incline to think that its 

 true affinity, but the wings of the rhachilla in G. cephalanthus are 

 orbicular, perhaps broader than in any other species of Eucyperus ; 

 so that Boeckeler has removed it to quite another place. Finally 

 we see, in common species like 0. longus and G, tegetum, very 

 great difference in the breadth of the wing in one species. — The 



