39 



block of plate iii, are sliown tubers of race 174, Khoai tiem froni 

 Saigoii, in which the fingers diverge, and race 78, a yam from 

 Manila, in which they do not. They are shown together for con- 

 trast, as it is easy to conceive transitions whereby the interval 

 between race 174, and races 38 and 72, might be bridged through 

 race 38. In race 28, which was received from the College of Agri- 

 culture, Los Banos, Philij^pine islands, as Tumuktok, the tuber ex- 

 hibits a tendency to divide into two which curve away from each 

 other, and in the angle Ijetwe'en them a series of smaller branches 

 may ai)pear. On ])late iv, lower block, one of the tubers photo- 

 gra])]ied shows down its side just such a series of smaller branches 

 as form in race 28. 



Kace 28 behaved slightly differently in 191G (vide ])late v of 

 the Bulletin Vol. I, double Xo. 11-12) ; and that this was so, is 

 interesting, for it suggests how consideral^le an influence the con- 

 ditions may have upon the form, — influences at present only to 

 be recognised as operative but to be analysed by further work. It 

 is well known that monstrosities of all kinds in plants appear fix- 

 able by selection, but the tendency to exhibit them is suppressed if 

 the plant is weakly or not under the best of conditions. So would 

 it be with regard to these yams ; and therefore if their characters 

 are to be brought out, they must be richly cultivated. 



This brings us to the conditions of the 1917 crop. An attempt 

 was made in 1917 to get more into the available ground by closer 

 planting than in 1916. Most of the sets were planted 2 feet by 3 

 feet (only the up growing yams 2 feet by 9 feet) whereas in 19 Ki they 

 had been 2 feet by ~) feet. This closer ])lanting while agriculturally 

 correct because tlie yield was hereiby increased, was not sound from 

 the botanical standpoint, as the jjlants competed with each other too 

 "much to develop quite freely and fully, the returns from each race 

 at the same time varying inconsistently from those got in 1916. 

 The following arc tables comparing the two years. 



