30 CHLORIS BOREALI-AMERICANA. 



with the admitted character of Gentianaceae is, unfortunately, one 

 which is placed in the foremost rank by Grisebach, namely, the aes- 

 tivation of the corolla. In Obolaria, the lobes of the corolla are 

 imbricated in the bud, instead of convolute, as in the true Gentia- 

 naceae, or induplicate, as in the Menyantheas.* On this account, it 

 might be proper to consider the genus as the representative of a 

 third group, of equal rank with the Mcnyanthese ; and in this form 

 it will accordingly be appended to the order Gentianaceae, in the 

 forthcoming portion of the Flora of North America, by Dr, Torrey 

 and myself. 



%* One or two mistakes Iiavo been committed in the analyses on the accompanying plate, 

 which were not observed in time for correction. In Fig. 3, the leaves of the calyx are wrongly 

 represented as decidedly distant from the base of the corolla, while, in fact, there is no such 

 manifest interval. In Fig, 4, the sinuses of the corolla (laid open) should be of equal depth, and 

 should extend to the insertion of the stamens ; the filaments, moreover, are rather too short. 



Tab. III. Obolaria Virginica, of the natural size. Fig. 1. Diagram il- 

 lustrating the ordinary disposition, &c., of the three flowers of the axillary 

 cluster. Fig. 2. Diagram of the occasional disposition of the same. In both, 

 the lower crescentic line represents a section of the subtending leaf ; the upper 

 circular one, a section of tlie axis ; the outer pair of the smaller crescentic 



* Perhaps there is a tendency at present to consider the characters drawn from 

 aestivation as more absolute and constant than they really are. Exceptional cases, 

 as well as variations in the same species, will be found by no means uncommon. 

 For example, although the sestivation of the petals is deemed to be universally con- 

 volute or twisted in Geranium and its allies, yet they are sometimes regularly 

 imbricated in Geranium maculatum, or, in a greater number of cases, while four of 

 the petals are convolute, the exterior one is wrapped around the others in the bud. 

 I have also observed this anomaly in G. Robertianum, G. sanguineum, and G. coUi- 

 num. So, likewise, the sestivation of the petals of Boykinia aconitifolia, NuU., is 

 convolute, or perhaps sometimes convolute with the outer petal imbricative, while 

 in other true Saxifragese the aestivation is regularly imbricated. 



