IX] CREEPING PLANTS 117 



la the Strawberry, and some Saxifrages (e.g. *S'. sar- 

 mentosa, S. flagellaris, &c.) and Houseleeks, the older 

 nodes, separated by long or short internodes of the 

 creeping shoots developed from the axils of leaves, do not 

 produce leaves, or produce only rudimentary scales, and the 

 terminal bud becomes anchored by roots to the ground, 

 while the thin connecting runner dies off sooner or later. 

 In some Houseleeks indeed the terminal bud breaks 

 away from its short runner as an independent structure : 

 it is clear that here we have a case where no difference in 

 principle exists between the bulbil and this terminal bud 

 of the runner — suppress the latter, and the separated 

 axillary bud is a bulbil. 



Where annual plants send out prostrate shoots — e.g. 

 species of Anagallis, Polygonum, Veronica, &c., they die 

 away at the close of the vegetative period and do not 

 serve to spread the plant : such shoots are merely de- 

 scribed as prostrate, or procumbe^tt, and if their terminal 

 portions curve upwards into more or less erect shoots they 

 are termed ascending (e.g. Ranunculus repens, many species 

 of Veronica and grasses). 



In some works on descriptive botany distinctive names 

 are given to special cases of the above creeping shoots, 

 e.g. trailing, procumbent, or prostrate, when they lie loose 

 along the ground but do not root at the internodes (Peri- 

 winkle, Knotgrass) ; repent or creeping, when the prostrate 

 shoot roots at the internodes, and dies off behind (Couch- 

 grass, Mint); the stolon, surmentum, or runner, has very 

 long and slender internodes, and some of the nodes not 

 rooting and with scale-leaves only (Strawberry) ; while the 

 short stolons of Houseleeks, Lobelia cardinalis, &c., with 

 their detachable terminal rosettes are termed offsets. But 

 it is not found easy in practice to maintain the distinctions 

 in all cases, for many trailing shoots will form roots at the 



