110 DR. HOOKER AND DR. THOMSON'S PR^CURSORES 



they have perennial rhizomes and annual branches ; a few are suf- 

 fruticose, having woody stems and annual branches. The axis 

 presents numerous modifications of habit, consistence, surface, and 

 extension. Some of the largest suffruticose species (J. Jurpia, 

 fruticosa, &c.) attain 5 feet in height — and perhaps much more — 

 resembling acanthaceous bushes in their half-shrubby character and 

 tumid nodes ; the scapigerous section has been already mentioned ; 

 another equally remarkable set of species, as regards their stems, 

 includes the succulent-jointed epiphytic peninsular species, I. Jer- 

 donice, viridijlora, &c. 



The morphology of the flower of Impatiens has occupied the 

 attention of many observers, who have taken different views of the 

 real nature of the pieces of the perianth ; we coincide with the 

 view taken by Eoeper*, Edgeworthf, and Henfrey]:, and which 

 we have ourselves confirmed from numerous observations on the 

 living species. There are three independent methods of obtaining 

 proof of this view ; and they all appear to us to give the same 

 result : they are the following : — ■ 



1. Seeking amongst the species for those with the fullest comple- 

 ment of organs, and ascertaining the position and value of the 

 supernumerary ones. — The existence of species with four lateral 

 sepals was first indicated by Edgeworth, who has pointed out that 

 they are figured in the plate of I. glanduligera (J. Boylei, Walp.) 

 in Eoyle's ' Illustrations,' who, moreover, detected them in other 

 species, I. amplexicaulis, &c., and first discerned their significance. 

 We may add I. longipes, urticifolia, leptoceras, var. r), sulcata, 

 sahrida, and arguta as also having four sepals, but not constantly. 



2. Comparison with a closely allied genus which presents a dif- 

 ferent modification of perianth, — as Hydroceras, which at once 

 explains Impatiens. 



3. Examination of the relative positions, &c., of the parts in 

 single and double flowers, with one another and with the axis — the 

 course pursued by Mr. Henfrey. 



All these methods lead to the same result, and appear to us to 

 prove that the flower of Impatiens consists of a calyx of three (or 

 more rarely five) sepals, of which the two supernumerary are 

 always smallest and next the axis ; the two next always green, and 

 small as compared with the petals and the odd one, or that away 

 from the axis, which is usually coloured and spurred. The corolla 

 consists of one erect, often bifid or emarginate petal next the axis, 



* Linnsea, ix. 921. t Iauu. Trans, xx. 37. 



X Linn. Joum. Bot. 



