AD TLOEAM IKDICAM (bALSAMINE^) . 109 



unable to avoid. The position of the bracts must in all cases be 

 referred to in order to determine the real condition of the inflo- 

 rescence, though not necessarily to find the place of the plant in 

 our system ; for we are obliged to place a species with two 1- 

 flowered axillary peduncles, bracteate at the base, in a different 

 section from another with a solitary 2-flowered peduncle, bracteate 

 at the ramification, however closely allied they may be in every 

 other respect. A more awkward fact still, is that both these 

 modifications may occur in the same species, sometimes in the 

 same specimen. 



An equally perplexing combination of discordant characters is 

 afforded by the ramification of the peduncle itself, the inflores- 

 cence being racemose, interruptedly so, whorled and corymbose, 

 or even umbellate through the reduction of the racemose or verti- 

 cillate inflorescence. Numerous examples of several of these 

 modifications in one species, or even specimen, may be found in 

 the groups SubverticillatcB, TTmhellatcB, and Bacemosce. 



10. The foliation may be wholly basal (radical), as in the Sea- 

 pigerce ; wholly opposite, as in most of the OppositifolicB ; wholly 

 alternate, as in most of the JRacemoscp. \ wholly (or almost) ver- 

 ticillate or ternate, as in I. Griffithii and I. Gardneriana ; opposite 

 below and alternate above, as in I. amplexicaulis, I. latifolia, and 

 I, Thomsoni ; opposite below and ternately verticiUate above, as in 

 I. flavida ; verticiUate in the middle and opposite or alternate 

 above and below, as in I. trilohata, I. salicifolia, and sometimes in 

 I. latifolia. I. multiflora and I. trilohata are very puzzling species, 

 being sometimes opposite-leaved throughout the whole individual, 

 at others as constantly alternate, and others having some of the 

 leaves ternate. The base (in form — acute, rounded, or cordate) and 

 petiolation of the leaf are extremely variable in individual species ; 

 of the former we may instance I. repens and I. scapiflora ; of the 

 latter, I. insignis, I. leptoceraSj and many others. 



For the most part the foliage is glabrous; it is sparsely hairy in 

 a good many, pubescent in a few, tomentose or hirsute in still fewer, 

 villous or glandular on the surface in none. The margin is often 

 serrate or duplicate-serrate, with gland-tipped lobules, and is armed 

 at the base and on the petiole with stipitate glands, which are 

 always very variable in number, position, and form in each species. 

 No species has stipules properly so called ; but many have large 

 glands, solitary or fascicled, sessile or stipitate, at the bases of the 

 petioles on the stem. 



11. The stems of Balsams are annual or perennial ; or more often 



