OF THE SEEDS OF BABRINGTONIA AND CAREYA. 51 



thorities, from whom all systematists have copied, and that there are 

 three distinct modes in which the structure has been viewed: — 



1. As an undivided exalbuminous embryo, which is the state- 

 ment of Blume, followed by Endlicher and Meisner. 



2. As an embryo in the axis of copious albumen. This view 

 originated with Gsertner, and was adopted by Hamilton, Eox- 

 burgh, Wight, and Lindley. 



3. As an exalbuminous embryo in two layers, a view first pro- 

 mulgated by Wight and Arnott, and doubtfully adopted by 

 Griffith. 



The manifest contradiction involved in these different modes of 

 describing the same parts in a seed of considerable size induced 

 me to take the first opportunity of examining ripe and germinating 

 seeds. This I have now been able to do for two seasons in the 

 Calcutta Botanic Garden, and the structure is so simple, that it 

 will require much less time to describe it than has been occupied 

 in the enumeration of the views of previous observers. 



An inspection of the ripe seed of JBarringtonia or Careya shows 

 at once that it is not perfectly homogeneous. A transverse section 

 of any part of the seed presents, as in Gsertner's plate of JBarring- 

 tonia and Wight's of Careya, two concentric layers, separated by 

 a ring of darker-coloured tissue, which has an organic connexion 

 with both layers. A longitudinal section, as is shown in Gsertner's, 

 Wight's, and Lindley's plates, as weU. as in those of Griffith, ex- 

 hibits the central body extending throughout the whole length of 

 the seed, and surrounded by the supposed albumen, from which it 

 is separated on either side by a narrow line of darker-coloured 

 tissue. The shape of the central body is dependent on the shape 

 of the seed, and therefore varies in different species of the two 

 genera ; but the relative position of the parts remains the same 

 in all. 



The microscope shows that both of these bodies consist of ordi- 

 nary cellular tissue full of starch-granules ; but that the separating 

 layer, which is in organic connexion with both, consists of a very 

 thin or almost single layer of delicate wood-cells (pleurenchyma) 

 intermixed with barred and true (unreliable) spiral vessels. 



The integuments of the seed are readily separable in Careya ; 

 they adhere somewhat firmly both to the fruit and the seed in 

 Barringtonia, but can be detached with a little care from the em- 

 bryo, most easily near the plumule. An examination of the sur- 

 face of the embryo before germination shows that, except two 

 minute and scarcely perceptible notches, first noticed by Griffith, 



E 2 



