394 Journal of the Asiatic Society of Bengal. [NS., XIV, 
stant enough to serve as distinctive marks and in fact the 
mass effect of the different shades is often quite striking. 
Pubescence.—Some types are almost glabrous while others 
are quite hairy: types which are light yellow in colour seem 
often to bear more hairs than darker coloured types. The 
character can be used in differentiating type from type. 
Seed Colour.—The seed may be of the following colours: 
White, grey or drab, greenish, brown, black. 
These colours have probably much significance in connec- 
tion with the quality of the oil produced. Howard states (4) 
that in India the black-seeded varieties are supposed to give 
the highest yields and the white-seeded the best quality of oll 
A white-seeded Burmese type named * Thadunbyu’ is usually 
said by cultivators to give the best quality of oil and this co- 
roborates the Indian opinion. The present investigations, how- 
ever, have not yet reached the stage when they can throw 
. 
light on this aspect of the question. 
3. CLASSIFICATION. 
colour of the flower, even although the variation in that - a 
were wide enough for classificatory purposes. Characters ® 
capsule, have a marked effect on yield, then these also se 
take equal rank with the branching habit. From the m oo 
trial point of view it is likely that seed colour will be of greet =a 
est importance. Hence as a tentative scheme the : 
classificatory scheme for the Burmese Sesamums is proposed. 
i Much branched, late, white seed. 
as 2» » .. coloured seed. yo 
B. Unbranched (less than two pairs of branches) ean 4 
white seed. 4 
BI. Unbranched, early, coloured seed. a 
at there 
_ The disadvantage of this scheme is of course prance 
18 continual gradation from little branched to much Jassifica” 
types, the division is accordingly arbitrary and the “Tt gives 
tion is avowedly more agricultural than botanical. he far: 
four main classes under each of which the types °” ve 
