381 



This race came from Assam just as did Xo. 162. But instead 

 of beiuo- a Khasia Hills race it was obtained at a forest-buried 

 stream-side village inhabited by Mikirs, where the depth of the 

 soil must have suited it admirably, and where wild animals abound 

 such as would destroy surface rooting races. 



The adjoining Khasia Hills furnish other similar races and 

 one was grown under the numljer 112, collected in 1911 for the 

 writer under the name of Phan Suri, and considered by the Khasias 

 living under Cherrapunji as the best of all yams. Like Xo. 186 

 it had pale yellow flesh. Xine plants yielded an average return of 

 4 lbs. 1 oz. (1833 grammes) or at the rate of nearly 4 tons 2 cwt. 

 per acre. 



, Again a not dissimilar race appeared among the plants of 

 Fijian origin, namely one called there " Boti " and numbered 120, 

 and another without a vernacular name numijered 110. The sur- 

 face of the tubers was rooty : the flesh was yellowish, with just a 

 very little magenta sap under the skin. The yield was on the 

 average 6 lbs. 9 oz. (2977 grammes) which works out at the rate 

 of nearly 6 tons 12 cwt. 



As regards minor characters, all the three races with these 

 elongated tubers possessing the tendency to branch, have similarly 

 shaped foliage, the auricles of the largest leaves being rounded : 

 but whereas the two Assam races possess no magenta sap, neither 

 in stem nor in tuber, the Fijian race. Xo. 120, has a little in the 

 tuber and in the frills of the stem. It alone is without prickles. 

 Tliat three of them have a yellowish flesh may prove of some in- 

 terest. 



The Fijian races, Xos. 156 and 118, which are represented in 

 the lower block of plate 1, differ from the races of the last groups 

 chiefly in length, and a tendency to thickening into a club. Simi- 

 lar to them are five other races now to be reported on, namely Xos. 

 108, 114, 116 and 136 from Fiji, and Xo. 54 from the Philippine 

 islands. 



If the reader will kindly turn to the plate he will see that Xo. 

 156 on comparison with Xo. 162 above, is not only shorter, but is 

 more covered with rootlets. Its flesh is white, magenta sap being 

 absent entirely. The average yield was 6 lbs. 5 oz. (2863 grammes) 

 or nearly 6 tons 7 cwt. per acre. 



Xo. 108 agreed in tuber exactly with Xo. 156, but yielded 

 more heavily — 9 lbs. 6 oz. or 4252 grammes, the return i)er acre 

 being almost 94 tons. 



Xo. 118 differed from Xo. 156 in possessing a little magenta 

 sap in a layer under the skin. Its average w^eight was 7 lbs. 12 oz. 

 (3515 grammes) or at the rate of 7 tons 154 cwt. per acre. 



Xo. 114 agrees very closely with the last, having the same 

 shaped tnl)er and similarly a little magenta sap under the skin. 

 Its averfige return was 8 lbs. 8 oz. (3854 grammes) or at the rate 

 of 8 tons 104 cwt. per acre, 



