38 Prof. Link on the Structure of the Orchidacese. 



among which one on each side might readily be supposed to 

 indicate a stamen on each side. The other canal, which is there 

 represented, originates from the confluence of the labellum with 

 the column, and whenever fertilization was artificially effected, I 

 found pollen- tubes in this canal also. Similar transverse sections 

 of the column of other Vandece always exhibit a quantity of vas- 

 cular bundles surrounding the stigmatic canal. It is clear there- 

 fore that there is still another part which surrounds the column, 

 and with the labellum represents a special organ, which must be 

 referred to the Linnsean nectary or to an accessory corolla (para- 

 corolla). This accessory corolla has two lips ; one, the upper lip, 

 blended with the column, the lower lip being the labellum. 



The comparison of the Orchidacese with the Alpiniacese lies 

 near, and indeed has been occasionally made, although mostly 

 superficially. The calyx of the Alpiniacese is spathaceously tri- 

 foliate, and, according to Lindley, corresponds to the calyx of 

 the Orchidacese : the corolla of the Alpiniacese always has two 

 divisions ; the outer tripartite envelope can only be compared, ac- 

 cording to Lindley, with the corolla of the Orchidacese, where 

 however that third leaflet is wanting, being absorbed, as it were, 

 into the labellum situated above it. There is nothing in the 

 flower of the Orchidacese corresponding to the inner division of 

 the flower of the Alpiniacese, unless the envelope of the column, 

 above spoken of, is taken into consideration. This, together with 

 the labellum, corresponds to the inner portion of the flower of 

 the Alpiniacese, in which there is always a well-marked labellum, 

 and very often, for instance in Hedychium and Globba, an upper 

 lip, which is merely not blended with the stamens and style as 

 in the Vandece. This upper lip is often wanting and the label- 

 lum exists alone, as in Alpinia, Zingiber and Kcempferia, just as 

 in our indigenous Ophrydece. The connecticulum is very much 

 expanded in the Alpiniacese, and so it is in the common anthers 

 of the Ophrydece, as well as in the calyptrate anthers of the 

 Vandece, and indeed in all Orchidacese the two anther-cells are 

 connected above by a membranous or fleshy portion, which may 

 be aptly named a connecticulum. 



As to the anthers, I will merely observe that the pollen of the 

 Vandece does not always lie naked upon the cellular body which 

 serves as its basis, but is inclosed in a delicate membrane of an- 

 gular parenchymatous cells, as I have distinctly seen in many, 

 particularly in Hunt ley a violacea. 



Cypripedium is not diandrous. The column divides into two 

 branches, each of which bears an anther-cell with two pollen- 

 masses. This division of the column alone distinguishes this 

 genus from the rest. Only one anther exists, but its chambers 

 are very much separated, as is usual in the Alpiniacese. My re- 



