Habitat. 



10i INDIAN FOREST MEMOIRS. 



but the author writes : " I can find no difference whatsoever between the specimens with and 

 those without the empty glume III." All specimens of the local plant examined have no 

 neuter glume and fall under (1) above. 



The description under (A) above which (with the exception of the characters in paren- 

 theses) refers exclusively to the local plant (i.e. to the form madagascariensis), growing 

 under varying conditions of soil-moisture, shows that this is a decidedly less robust plant 

 than the species as understood by Prain and in the F. B. I. which also includes the form 

 Zollingeri ( = Arundo Reynaudiana Kunth). 



The latter also has a different area of distribution and although starved specimens of 

 both these plants exist which, especially in herbaria, are likely to mask the differences 

 between them, it seems probable that they will prove to be distinct species. The writer, 

 however, has at present no knowledge of the E. Himalayan plant in the field and pending 

 the results of further study it seems advisable to deal with these forms as varieties as 

 follows : — 



(1) madagascariensis proper : Glume III similar to IV and fertile. 



N. W. tropical Himalaya and Sub-Himalayan tract from Punjab to Kumaon, 

 Africa, Madagascar. 



(2) Reynaudiana : Glumes III and IV heteromorphous, III empty. 

 E. Himalaya, Assam, Bengal, Burma, China, Java. 



Specimens kindly sent me by Dr. Stapf, of the African plant from N. Madagascar 

 (Revd. R. Baron, 6353) with spikelets 0-31 to 0-34 in. long, appear to be identical with those 

 of the N. W. Himalayan form. 



C. — Biological and (Ecological Notes. 



83. This plant is sometimes associated with 

 Aristida cyanantha on gravel and boulders, but as a rule only becomes dominant where 

 there is a considerable admixture of sand, and where there is still insufficient moisture for 

 a vigorous development of Saccharum Munja. 



With it are often found the xerophilous form of Saccharum syontaneum and S. Munja 

 and also Andropogon monticola. 



In many respects this grass closely resembles Aristida cyanantha. Like the latter the 

 present species, in its usual local habitat, shows a well-marked resting season during 

 December — January. 



It however also grows vigorously on moist loam under cultivation and then shows a 

 less sharply defined period of rest. 



Plants taken from sand at once established themselves strongly in loam and the species 

 is therefore apparently not so sensitive as A ristida in respect of soil aeration. 



Isolated plants are occasionally met with growing well in shady Sal forest and it 

 probably has a stronger capacity for withstanding shade than Aristida. 



[ 104 j 



