10 Dr. Grisebach on the Germination 



bent, which I did not see in any other plant of the family. But 

 another difference is afforded by the testa of the ripened seed, 

 which is covered by a thick and somewhat woolly epidermis 

 in all Menyanthidece, while the testa of the Gentianece is quite 

 simple : the epidermis of the former is often muricated, a cha- 

 racter which seems to be very constant, and may be employed 

 for specific distinctions. If we consider that the Menyanthidece 

 are aquatic, many of them even floating plants, we may ascribe 

 the presence of the epidermis of their seed to their habitat ; for 

 a similar one has not been observed in the Gentianece, nor can 

 we consider this lobe a distinctive character, as it may disap- 

 pear, when either one of the Gentianece is detected growing in 

 water, or one of the Menyanthidece on dry ground. Finally, the 

 placentae rise, in Menyanthes trifoliata, from the middle of the 

 valves, which seems to result from an union of them with the 

 interior surface of the endocarpium, and which occurs also in 

 many Gentianece ; the seeds are commonly much larger, but 

 in Frasera they are equally large : in the latter genus their 

 number is subdefinite, and in Villarsia ovata it is reduced, 

 lastly, to a single seed on each side. Finally the induplicative 

 aestivation of the corolla affords a good character for the Meny- 

 anthidece as a peculiar group of the family, being dextrorsum 

 contorta in all Gentianece ; but the transition of these charac- 

 ters takes place in those Gentianece which have a plica between 

 the lobes of the corolla, these plicae having also an induplicate 

 aestivation ; and, as in G. Andrewsii, the lobes disappear wholly 

 by abortion : the corolla by the plicae remaining affords the 

 same aestivation as in the Menyanthidece. So far with regard to 

 the differences of these groups in their reproductive organs, for 

 I know of no more than these, and I find besides strong ana- 

 logies, as for instance, in the direction of the unequal sepalum 

 being remote from the axis (fig. 7-) . As to what concerns the dis- 

 parities of vegetation, these depend mostly on their aquatic or 

 terrestrial station, viz., the development of a rhizoma, and ac- 

 cordingly of alternate leaves (which are repeated in Swertia), 

 the development of long petioles, so that the leaf may arrive 

 at the surface of the water, and the verticillae of leaves in Meny- 

 anthes trifoliata being remote from its stem for the same 

 purpose, &c. I think an accurate examination of all these 



