181< LETTER XLIV. 



they come in contact projecting into five whitish angles, 

 adhering firmly hy their faces to a pentagonal, flattish, 

 fleshy, red and green table, which they surround (fig. 

 7. «.)» ^^^ having each a whitish, membranous termina- 

 tion, which curves over the table aforesaid {fig. 5. e.). 

 In the next place, carefully remove two of the anthers, 

 turning them on their backs (fig. 6.) ; you will find 

 that each is two-celled, and that the pollen of the con- 

 tiguous cells of two different anthers, forms two orange- 

 coloured bags (fig. 6. a. and fig. 9.), which are very 

 loose in their cells, and adhere to a blackish, oval 

 gland, that belongs to the angle of the table aforesaid 

 (fig. 7. b. 6. b. & 9. b.) ; so that when you open the 

 anthers, you see the bags dangling from the gland like 

 a pair of yellow pouches (fig. 90- 



After all this apparatus is removed (as at fig. 70» 

 you have a view of the pistil, consisting of two 

 ovaries placed in close contact, and each containing 

 a large, fleshy placenta, covered with ovules (fig. 8.). 

 To each ovary is a single style, which is placed parallel 

 and in contact with that of its neighbour, without 

 uniting to it (fig. 7- c.). The styles are held together 

 by the fleshy five-cornered table that surmounts them 

 ( fig. 7. ci.)t and which stands in the place of a stigma, 

 without exactly being one ; for the influence of the 

 pollen is not communicated to the ovules through its 

 tissue, as in true stigmas, but somewhere about the 

 point where the style and the table join (fig. 7- ^O* 



When the corolla and stamens have fallen off, the 

 table and styles give way, the two ovaries diverge, and 

 if both of them continue to grow, you will find, when 



