HOLE : SOME INDIAN GRASSES AND THEIR CECOLOGY. 83 



of abnormal panicles of Saccharum Narenga which owe their peculiarities of a fistular 

 calm and persistent panicle branches to some injury similar to that described above. This 

 species therefore cannot be maintained. 



It is noticeable that the injurious action of the insect, or fungus, or both, described 

 above can to some extent be imitated artificially. On a plant growing in the experimental 

 garden at Dehra Dun one of the flowering culms was sharply bent over and left hanging on 

 the plant on October 20th, 1909. It was thought that this might interfere to some extent 

 with the passage of water and food materials to the panicle and thus more or less produce 

 the results mentioned above. The panicle which was treated in this way did show the 

 persistent branches and general appearance of those already described but the culm did not 

 become fistular, probably owing to the central tissue having been fully developed when the 

 culm was bent. 



Plate XXIII, fig. 2, shows the panicle treated as above described and three other normal 

 panicles, all four of these having been picked from the same plant on November 20th, 1909. 

 The panicle from the bent culm clearly shows the persistent branches. 



It must be noted that culms are sometimes found in the forest which have been more 

 or less damaged by insects, or fungi, or both, but which do not show the abnormal characters 

 noted above, and the extent to which the culm and panicle are affected appears to depend 

 on the state of development of the culm when it is attacked and on the severity of the attack. 

 In no case, however, has an abnormal flowering culm, exhibiting the characters described 

 above, been found which had not been more or less severely damaged, while both abnormal 

 and normal culms have frequently been collected from the same plant. 



It must also be noted that the flowering culms of this species and also those of 

 Saccharum Munja and Erianthus Ravenna? are frequently attacked by larvae which hollow 

 out the culms and this also often results in the premature drying of the panicle but this 

 kind of damage obviously differs from that described above in which no obvious excreta 

 or other signs of larval damage are seen. Hackel divides the present species into two 

 varieties as below : — 



a genuinum Glumes I and II of sessile spikelet obtuse, glabrous dorsally. 



I sub-enerved with the exception of the keel nerves. 



Sessile spikelet 2-5 to 3 mm. (0-1"— 0-12") long. 



Tropical Himalaya. 



/3 Khasianum Glumes I and II of sessile spikelet sub-acute, dorsally hairy below. 



I faintly 3 nerv< d in addition to the keel nerves. 



Spikelets larger 3-5 mm. (0-14") long. Khasi Hills. 



The writer finds it impossible to follow this classification. In the first olace it must be 

 noted that in dried specimens the hairs easily break off from the hard coriaceous glumes, 

 especially in mature spikelets, and leave no obvious scar. It is not therefore always easy 

 to determine the degree of hairiness of the glumes from herbarium specimens. In all the 

 fresh flowering spikelets collected by the writer in Dehra Dun, however, there is an obvious 



[ 83 ]! 



