388 



Since the preceding part was written I have again examined and 

 compared all the specimens of Gentiana at my command, that can be 

 refen-ed to either of the two forms. This examination has led to the 

 following conclusions. 



1. That no character can be derived from either the positive or re- 

 lative size of either plants or flowers ; both forms being exceedingly 

 and equally variable in this respect. The specimen from near Streat- 

 ley is much larger, more branched, and with many more flowers than 

 the specimens of G. Amarella which accompanied it; the reverse of 

 these conditions obtained in the plants mentioned in the extract from 

 the ' Gardeners' Chronicle' (Phytol. 381). 



2. That no dependence can be placed on the relative proportions 

 of the leaves, branches, and internodes ; these proportions frequently 

 varying even in the same individual. In the Streatley specimen some 

 of the branches are shorter, others are longer, than the internodes of 

 the stem, while the leaves are shorter. The same remarks will equally 

 apply to large-flowered continental plants, and to British specimens 

 of G. Amarella. 



3. That in form both the leaves and calyx-lobes are variable ; and 

 the proportion borne by the latter to the tube of the calyx difiers in 

 the same^specimen. 



4. That although the tube of the corolla may, in some specimens, 

 be correctly described as cylindrical, and in others as gradually wid- 

 ening upwards, yet intermediate forms are extremely common. And 

 Grisebach, in his descriptions, calls them both " obconical," as Mr. 

 Brown has already remarked, (Phytol. 320). 



5. That with respect to the stalk [stipes) of the ovary, if by that 

 term is to be understood a stalk separate and distinct fi'om the sub- 

 stance of the ovary supported by it, similar to that represented in tab. 

 236, ' English Botany,' then such a stalk I have not been able to find 

 either in the Berkshire plant before mentioned, in continental speci- 

 mens, or in any others that can possibly be referred to G. geimanica. 



6. That if, on the contrary, by the term stipes we are to understand 

 such a gradual diminution of size in the base of the ovary itself, as 

 would probably be produced by the shrinking of that part, as sug- 

 gested by Mr. Brown, then such a stalk I find to be more or less evi- 

 dent in every specimen that I have examined, whether referrible to 

 G. Amarella or G. germanica. 



7. Moreover, in undoubted specimens of G. Amarella, gathered 

 rather late in the season, I find some flowers with mature capsules, 

 which contain ripe seeds, and are generally perfectly sessile ; while in 



