PHANEROGAMIA : FLOWERS. 389 



seed-buds, which especially prevails in the cases where the spermophore 

 is central, or belongs to the inferior germinal cavity ; if we here put out 

 of view the metamorphosis of the fundamental organ, and continue the 

 arrangement of the carpellary leaves in equal-membered alternating 

 circles, we obtain, with two carpels four-ranked, with three six-ranked 

 leaves, therefore four or six rows of axillary buds. By the conversion 

 of the axis into the spermophore the regular rows of buds are pressed 

 nearer together in pairs. If we examine, in this respect, Tillandsia 

 amcena, we find really six spermophores, which are approximated in 

 pairs, but neither so morphologically nor anatomically connected as to 

 justify us in regarding them as three two-leaved organs. 



Where, however, the spermophores are to be regarded as lateral 

 branches of the floral axis, whether in germens formed of carpels or in the 

 stem-pistil, we must always look upon them as flat shoots folded together, 

 which then, somewhat like the flat shoots of Phyllanthus, bear two 

 rows of buds. 



On the structure of the spermophore I have no additional remarks to 

 offer, since no especially striking conditions are known to me. 



c. Of the Seed-bud (Ovule). 



161. Every seed-bud (gemmuld) appears at its first origin as a 

 email roundish papilla at the extremity of an axis (terminal shoot) 

 within the flower ; as such it is an erect, straight seed-bud (aemmula 

 erecta, atropa). It consists simply of the nucleus (nucleus, chorion 

 [Malpighi], perisperma [Treviranus], lamande [Brongniart], tercine 

 [Mir Del]), without special integument (nucleus nudus). On this 

 seed-bud are to be distinguished the base (when it does not gradually 

 pass into the axis of which it forms the extremity) at the point of 

 attachment of the seed-bud (hilus, umbilicus), and the apex as 

 nuclear papilla (mammilla nuclei, mamelon tf impregnation [Brongn.] ). 

 The seed-bud seldom persists in this simple condition, as in the 

 LoranthacecK. It usually changes, partly by the formation of the 

 integuments of the bud, and partly by the peculiarity of its modes 

 of development, which may in general be termed curvatures. 



At a greater or less distance beneath the apex of the seed-bud 

 there arises simultaneously on the whole circumference a circular 

 fold, which gradually rises over the nucleus and closes in it above, 

 leaving only a small orifice. If this development is permanent (as 

 in Taxus and the Piperacece), this envelope is termed a simple bud- 

 integument (integumentum simplex) ; the upper opening is called the 

 micropyle, and the part where the bud-integument and nucleus are 

 confluent is called the base of the bud (chalaza). The point of 

 attachment is here precisely defined, as in the former case. Below 

 the first circular fold, of which we have spoken, a second is often 

 formed, which encases the first as the first encases the seed-bud. 

 The first is then styled the inner bud- integument (integumentum 

 primum internum, membrana internet [Robert Brown], teamen 

 [Brongniart], secondine [Mirbel]). The other is the second or outer 



c c 3 



