163 



grown berries of 5. lanceolata the elastic coat is even more readily 

 detected than in the ripe ones of S. mtrictdata ; the dark pulp, 

 however, appears as a layer of firm, green, fleshy tissue, thick at 

 the base and thinly enveloping the entire mass of albumen, which 

 is still white exteriorly and yielding (though tough) in texture- 

 Under careful dissection the funiculus sometimes separates at its 

 base and remains attached to the seed — projecting from the elas- 

 tic coat as a minute thread-form body, perhaps a fiftieth o{ an 

 inch in length. More often, in the ripe berry, it gives way at the 

 other extremity and remains attached to the angle of the cell, 

 presenting at the free end a saucer-like, sub-circular expansion, 

 by which it was attached to the hilum. 



It is absolutely certain that the elastic exterior coat and the 

 enclosed pulp layer are borne upon the funiculus and must be 

 regarded as integral parts of the seed, or as adjuncts to it, and 

 not as any part of the pericarp. It may be questioned whether 

 the elastic coat is testa or aril. I am myself strongly inclined to 

 believe that it is merely the developed outer coat of the ovule 

 and therefore the testa of the seed. Accepting this view, the 

 brownish closely adherent covering of the albumen (testa of A. 

 de Candolle) is really the tegmen. This structure is simple 

 enough, and is paralleled in numerous bitunicate seeds. The 

 fleshy and subsequently pulpy intermediate layer remains, how- 

 ever, a curious anomaly. Gray appears to hold (Struc. Bot, ed. 

 'S79, p. 277) that the coats of the ovule are never separated. 

 They clearly seem to be in this case, and yet to no obvious pur- 

 pose. It is a suggestive fact that in the ripe fruit the pulp 

 enclosed within the testa (if testa it is) presents decidedly an 

 appearance of decomposition rather than o{ wholesome maturity. 

 Apparently from some distorted or at least unusual impulse, the 

 outer coat of the ovule begins to develop some little distance 

 below the inner. Nature, it seems, abhorring the vacuum thus 

 formed, fills the gap with tissue, which merely serves as a pack- 



wnlers carefully consulted on lias point may be named Tournefort, Linnaeus, Gaert- 

 "er, Michaux, Nuttall, Torrey, Gray, Chapman, Wood, Kunth, Grisebach, A. dc 

 Candolle, Bentham and Hooker and Engler and PrantL The hilum seems to be 

 unmentioned, except by Kunth (Enumeratio) and Grisebach (Mart. FI. Bras.); only 

 Gaertner notices the tubercle at the apex. Strangely enough, Gartner figures the 

 ":mbryo at the wrong end of the seed ! 



