66 



JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



displaced the species now cultivated in Northern India (Indigofera 

 sumatrana Gaertn.). 



During the first few years all attempts made to grow this plant in 

 Behar were signally unsuccessful. Even when the seed was sown on the 

 abundant moisture, and during the high temperatures, prevalent during the 

 rains, germination was poor and gave rise to a scattered irregular crop. 



It soon became evident that Indigofera arrecta affords yet another 

 example of a leguminous plant, the seed of which is characterised by the 

 possession of a " hard coat." The recognition of this fact has removed 

 all those doubts as to the possibility of growing this plant in Behar which 

 the previous failures had excited. The only remaining problem is to 

 obtain a cheap and efficient method of " treating" (slightly scratching the 

 seed-coat) seed in bulk. It is hoped that the remedy has been found in 

 the Svalof machine,* devised more especially for treating clover-seed, in 

 which the same defect occurs. 



The extent to which germination is injuriously affected is shown by 

 the following figures : — 



I. arrecta (Natal), 



Seed from plant grown in Behar (1902) 

 „ „ „ ,, „ (1903) 

 Seed from plant grown in North- West Provinces (1902) 

 Seed imported from Natal 



Per cent, 

 germination. 

 8 

 6 

 10 



No opportunity has, so far, occurred for examining seed from Java, 

 where, as far as information goes, no difficulty is experienced with the 

 germination. The seed of " Java " plant, grown for one season in Behar, 

 showed a germination percentage of seven in a test lasting seven days. 



With the exception of the last, the above figures represent the per- 

 centage number of seeds which have germinated in five days. This 

 number will, of course, be slightly increased by continuing the experiment 

 for a further period. Five days, however, are sufficient to cover the period 

 of germination of seed which has been treated in an efficient manner. 

 Thus : 



Day 



2 



3 



4 



5 



6 



Total 



Per cent. 

 Germination 











Germinated 



Hard 



Untreated seed 

 Treated seed . 



1 



16 



1 



38 



2 



16 



3 

 9 



1 



0 



8 

 79 



150 



30 



5 

 72 



That the above figures for treated seed are not a true measure of the 

 " good " seed in the sample is shown by the following table, the figures in 

 which are obtained from a second sample of the same seed : 



Germinated 

 Second day 95 

 Third day 65 



Fourth day 7 

 Fifth day — 



Sample of 230 seeds. 



Swollen Bad 



— — "1 unswollen 



12 0 j again treated and tested. 



Germinated Swollen Bad 



5 — 25 29 — 



— 5 25 — 4 



54 



* A short account of this machine occurs in a Report on the Scandinavian seed- 

 control stations, in the Experiment Station Record, U.S. Dept. of Agr., vol. x., 1898. 



