'HARD SEED" IN INDIGOFERA ARRECTA HOCHST. 67 



From this table it is clear that the true percentage of "good" seed is 

 96 (221 seeds in 230), and the percentage of " bad " seed is 4. It is also 

 clear that the failure to obtain germination in 39 cases (see previous table) 

 was due to ineffective treatment. In the latter table a similar failure 

 occurred in 59 cases, making the apparent germination percentage 74. 



No serious effort has- been made to ascertain definitely the cause of 

 this phenomenon. An ash determination alone has been made and the 

 figures for I. arrecta compared with those for I. sumatmna, the seed of 

 both of which species was obtained from plant growing in the same 

 locality. 



I. arrecta I. sumatrana 



Testa taken 1-617 g 1-334 g 



Weight of ash after ignition . . 0*022 g 0-027 g 



Percentage ash . . . 1-36 2*02 



Ash not deliquescent Ash markedly deliquescent. 



These figures seem to indicate that the phenomenon is not the result 

 of the occurrence of a large amount of mineral matter in the testa,* and 

 the second explanation, which traces the resistance to the organic, and 

 not the inorganic, substances composing the testa, seems more probable. 

 In the outer layers of this, a body or bodies exhibiting a strong repulsive 

 attitude to water are supposed to be present. 



FIGURE. 



In left foreground. — Crop of I. arrecta (' Natal ' var.) seed, untreated, and sown 

 6 seers (12 lbs.) per bigha (| acre). 



In right foreground. — Crop of I. arrecta (' Natal ' var.) seed, treated and sown 

 4 seers (8 lbs.) per bigha. 



Seed was broad -casted, and hence the line of demarcation is neither straight nor 

 very definite. 



* Vide Agricultural Botany, J. Perciv 1, p. 623. 



