16 



ON HYBRIDIZATION AMONGST VEGETABLES. 



calyx somewhat unequal. The fact is not so. The two upper 

 lateral segments in Gloxinia speciosa and hirsuta have a dispo- 

 sition to be smaller than the others. The calyx of the order 

 consists of one upper segment and two pair of laterals, and there 

 is often a little and not very certain difference between the two 

 pair. Their relative size is manifestly unsteady. 2. Corolla 

 protuberant on one side only of the base. It will be found that 

 the protuberance varies in the species according to the greater 

 or less depression of the corolla. In Gloxinia speciosa it is bent 

 downwards so rigidly, that the base can only swell upwards. In 

 Gloxinia hirsuta the corolla is not depressed and the base is very 

 different. Proceeding to Achimenes, we find an assertion utterly 

 unfounded, that the anthers are separate, being united in the 

 two former genera. I have before me the anthers of Achimenes 

 coccinea, pedunculata, hirsuta, &c. all as closely attached to- 

 gether as those of any Gesnera. I have also seen a few flowers 

 in which they were separate, but I believe in a barren and im- 

 perfect state, and I find them more ready to part when they de- 

 cay. Having disposed of those false facts, let us see what dis- 

 tinctions remain. Gesnera, corolla tubular. Gloxinia, funnel- 

 shaped, or somewhat bell-shaped, inflated in the middle. Achi- 

 menes, tubular and funnel-shaped. Those differences, if cor- 

 rectly stated, would only show that there is a variability in the 

 swelling of the tube in different species, as there is in a much 

 more decisive manner amongst the various species of tropical 

 Convolvulaceae in the genus or section Pharbitis ; but those facts 

 are also incorrect. The corolla of Gloxinia hirsuta is not in- 

 flated, but nearly cylindrical, with furrows ; on the other hand, 

 that of Gesnera zebrina and Geroldtiana is immoderately in- 

 flated, and nearly similar in form to that of Achimenes pedun- 

 culata and hirsuta. What remains? Nothing between Ges- 

 nera and Gloxinia ; a ring round the ovary to Achimenes. I 

 can however point out one, which has not been noticed, of con- 

 siderable importance between the fruit of Gloxinia and Gesnera 

 as applicable to Gloxinia speciosa and hirsuta. The fruit erect. 

 Gesnera, the fruit sub-horizontal, with a beak curved a little 

 upwards and a different dehiscence. Again: Gesnera, calyx 

 ad pressed. Achimenes, calyx patent. No person can compare 

 Gesnera zebrina and Geroldtiana with Achimenes pedunculata 

 and hirsuta, and not see that they are much more closely allied 

 to the latter than to Gesnera faucialis and bulbosa, and others 

 which have the upper portion of the limb prominent and in- 

 curved, while that of the former is short, two-lobed, and recurved : 

 but the calyx of Gesnera rutila is neither patent nor adpressed ; 

 the calyx of Gesnera Geroldtiana has the upper lobe of the calyx 

 not adpressed, while that of faucialis is closely adpressed; there- 



