222 



CLASRIFIOATIOX OF RTCK 



Provinces rices as the nninbor of pi'jmary ])ranches is extremely 

 variable. Thus in • Meqlial ' ( Raipui) tlie average nuiu])ei' of branches 

 was 10 with a range from 6 to 14 in 83 specimens examined ; 'Jiro 

 (Narsinghpur) had an average of 16 with a range from 13 to 21 in 61 

 specimens examined ; 'Jhamul Meli (Raipur) had an average number 

 of 11 with a range from 8 to 15 in 105 specimens examined. These 

 three examples, taken from 64 medium rices examined, sliow how 

 unsatisfactory this character is as a basis for classification. A range 

 of eight is by no means uncommon, the ordinary range being six, and 

 a range of less than five the exception. Two further examples 

 will show that in certain cases the average number of prinmry 

 branches does not even represent the number of greatest frequency. 



Rice. 



Laxmibhog 

 Raipur 



Average number 14; number of greatest frequency 1"». 



Ari Motan 

 Kaipur 



Average number 9 ; number of greatest freq 



Spikdef. — The spikelets are articulated on their pedicels which 

 may be long or short. The pedicel is enlarged annulnily nt the top. 

 the two sides l)eing distinctly oblique, that is, one side is liiglier 

 than the other. Just below the enlargement the pedicel is bent or 

 contorted. In many rices the facet(15) or aiumhif (Milingcinent 

 is expanded so as to appear scale-Hke. Hence (V)ok(l()) in (h\sciib- 

 ing O.coarctata gives the number of glumes as 5, the spikelet being 

 jointed above the lowest pair. The size of the facets is of a certain 

 amount of value in distinguishing the varieties. 



