Chances of Life 249 



their growing than others. For with some, it is abso- 

 lutely necessary that they should be buried before they 

 can even begin to germinate; while others, though they 

 may be able to germinate without help or preparation of 

 any kind, find it by no means easy to do more than make 

 the first start. 



With some, germination is an easy matter enough, all 

 that they need being a little moisture. Such is the mus- 

 tard-seed, which will sprout, and even grow for a time, 

 on any damp surface; a piece of flannel, or even the out- 

 side of a porous earthenware jar will do, if only this is 

 kept filled with water. Other seeds begin to shoot even 

 before they leave the parent-plant. The seeds of a species 

 of convolvulus put forth quite large leaves before they 

 burst the pod; and in hot climates, the seed of the water- 

 melon sometimes grows in like manner, within the fruit. 



Brazil-nuts also begin sprouting before the hard outer 

 case in which they are inclosed shows any sign of decay; 

 but though they may all sprout, only one of the whole 

 number seems to have a chance of doing more under 

 ordinary circumstances. There are from twelve to fifteen 

 nuts — or, strictly speaking, seeds — in each ovary or case, 

 which is filled with the matted roots sent out by one and 

 all. At the lower end, where the fruit was attached to 

 the stalk, there is a small opening, and the fortunate indi- 

 vidual which gains possession of this exit may eventually 

 burst the case with its roots, and so make its way into the 

 soil. But the shell of the case is extremely hard, and so 

 far as has been observed, it is not often that even one 

 single plant succeeds in freeing itself. But then, on the 

 other hand, if the shell were less hard, none would prob- 

 ably ever escape the hosts of animals ready to devour 



