238 Mr. C. C. Babington on some species o/Epilobium. 



In addition to the characters used in this arrangement, the 

 following points deserve notice. 



1. The stem in some of the plants rises erect directly from a 

 fibrous root, and usually produces lateral branches from the 

 axils of its lowest leaves so as to take a rather csespitose form. 

 This primary stem appears always to be erect, but the lateral 

 stems or branches are usually procumbent at their base and fre- 

 quently produce roots there, although throughout the greater 

 part of their length they are erect or ascending. When the 

 plants grow in water, or in very wet places, these adventitious 

 roots are sometimes produced from the lower joinings of the 

 upright primary stem, and the procumbent part of the branches 

 is very long : if in this case a branch is carelessly pulled up, the 

 plant may easily be supposed to have a cordlike base, when its 

 real structure is very different. Towards the end of the summer, 

 or in the autumn, these csespitose species usually produce from 

 close to the base of their stem very short flowerless shoots 

 having their joints so much contracted that the leaves lie closely 

 upon each other, and a rosette or rose-shaped tuft is formed. 

 The original plant does not survive the winter, but in the 

 ensuing spring the place which it occupied is more or less sur- 

 rounded by a cluster of new csespitose individuals resulting 

 from the rosettes of the preceding autumn ; each rosette pro- 

 ducing from its terminal bud a new primary stem, and from 

 some of its axils a few lateral stems. 



In other plants, thick long stoles with distant leaves take the 

 place of the rosettes. It is only at the end of these stoles that 

 the least trace of the close arrangement of leaves forming the 

 rosettes is to be found, nor is it always seen even there. These 

 long stoles root and live through the winter, and their remains 

 when attached to the base of the stem of the succeeding year may 

 be taken for the chordorhizal structure if the stem fails to pro- 

 duce lateral stems from its lower axils. The character derived 

 from the chordorhizal base is not therefore wholly to be trusted, 

 although Fries has confidence in it. 



2. Another habit is that in which there is no trace of the 

 csespitose mode of growth, but in its place there is a prostrate 

 slender stem producing many adventitious roots, and turning 

 upwards at the end so as to form the upright stem of the plant. 

 If branches are at all produced from the lower part of the stem 

 they are placed at some distance from each other, or in distant 

 pairs, for the joints are long. Most of these species throw out 

 from many of their lower joinings stoles furnished with long 

 joints and pairs of very small leaves, and end in a sort of bulb, 

 the scales of which are rather fleshy with their tipper epidermis 

 loose. These bulbs become detached in the winter, by the decay 



