C. A. BARBER 135 



But a mimber of forms were not so easily arranged, and the " unclassified 

 list " steadily increased, without any light being obtained as to their relation- 

 ships to the groups already collected and isolated. It almost seemed as if 

 there were a large number of isolated forms growing in all parts of the country 

 without any visible genetic connection. Some, such as Barahi from the 

 Central Provinces, were as small as the thin and meagre varieties of the Punjab, 

 while others, such as Naanal in Madras and Mojorah in Assam, were with some 

 difficulty separated from the more juicy canes such as Puri of Bihar and Bonfa 

 of Madras, which appeared to have been introduced at some time, but, after 

 long growth in the country, had apparently decreased in size and vigour. In 

 fact, the rigid separation of exotic from indigenous canes has sometimes been 

 far from easy, and many changes in the unclassified list have been made in 

 consequence. Until a more detailed examination can be made, a certain 

 immber of forms will be under doubt, as it does not at first sight seem possible 

 to decide whether such varieties as Magh in Assam and Vendamvl'hi in Bengal 

 are in reality introduced or country canes. But the characters of the thick 

 and thin canes, such as their mode of growth and tillering, their size and the 

 relative hardness of their rind, the amount of fibre and juice and their resistance 

 to disease, make the separation of the two classes in general easy enough, and 

 this is emphasized by the character of such seedlings as have been raised 

 from them, in that the parental differences are often exaggerated in the 

 seedlings. 



Two varieties placed amojig the unclassified indigenoMs Indian canes, 

 still being grown occasionally in the midst of the thick canes of South India, 

 received special attention, because they were more familiar, and we had raised 

 seedlings from them. They were seen to differ markedly in two respects. 

 Ganda Cheni from Mysore had bending or nodding leaf ends and well developed 

 circlets of hairs at the nodes, while Naanal from Tanjore had more or less 

 erect-tipped leaves and was practically devoid of the circlets of hairs at the 

 nodes. In another place, habit, as here exemplified by the character of the 

 leaf ends, has been regarded as important in classification, whereas it has been 

 suggested that the presence of the circlet of hairs is a primitive character of 

 fundamental importance. These differences, therefore, seemed to indicate that 

 these two canes, hke one another in many respects, might belong to entirely 

 different groups. But there was no time available for classification studies 

 until the seedling work had been placed on a firm foundation. The two varie- 

 ties were obviously to be regarded as indigenous canes, when compared with 

 the thick ones of South India, but their systematic position remained a 

 mastery. 



