116 



tree and become established in the soil, the primary root axis 

 does not, at least in many cases, develop. In its stead, how- 

 ever, one, or usually two, secondary roots (Fig. 3) are formed 

 in the usual manner, i. e., laterally, very close to the end. 

 As a result of the mechanical relations of the tissues, these 

 are from the start forced to grow parallel to the chief axis, 

 and thus take the place of the chief root. That this substi- 

 tution of lateral hypocotyledonary (adventitious) roots ac- 

 tually takes place can be shown by the intentional destruc- 

 tion of the chief root, which, as above stated, sometimes takes 

 place by withering. The same thing may be induced higher 

 up in the hypocotyl by removing the lower end by a trans- 

 verse cut. Lateral injuries of various forms, even when they ex- 

 tended as far as the central cylinder, did not stimulate the forma- 

 tion of new roots, nor does this occur, excepting at the extreme 

 lower end after longitudinal splitting of the hypocotyl. The 

 growth of these lateral roots is considerably slower when the end 

 is removed, from which it appears that the tissues are the less 

 able to form roots, the further the point of injury is from the 

 original radicle. A month to six weeks may elapse before the 

 fundaments of these new roots may be readily seen, and the 

 general development of the seedling is correspondingly slow. 

 The plumule is, however, often well developed before the seedling 

 becomes separated from the tree (Fig. 1). 



A very interesting case of vivipary is recently reported by Dr. 

 O. Stapf* to occur in one of the tropical grasses {Melocanna 

 bambusoides Trin.) of the forests of Bengal. In this plant the 

 endosperm (presumably in the mature fruit) is lacking, while the 

 testa and pericarp are specialized to form nutrient tissues. The 

 scutellum is very considerably enlarged and occupies the space 

 otherwise filled by the endosperm. The scutellum, which is 

 richly supplied with vascular tissues, acts, during germination, 

 upon the pericarp in a manner analogous to its action in other 

 grasses on the endosperm. Dr. Stapf believes that certain other 

 genera (Melocalam?(S and Ochlandrd) offer similar conditions. 



Dr. J. K. Small has reported vivipary to occur in Tillaudsia 



* Nature, 65 : 548. 10 Ap. 1902. 



