62 



THE JOURNAL OF BOTANY 



S. sagino'ides var. macrocarim. 



On the average, about 1*6 

 mm. long and 2-25 broad. 



Usually erect (or suberect) 

 in fruit, rarely spreading. 



On the average, about 1'5 

 mm. long and 2*0 broad.'-' 



On the average, about 5-8 

 mm. long, and about 1'3-1"9 

 times as long as the calyx. 



S. sagino'ides var. iyinca. 



Sepals 



On the average, about 1*3 

 mm. long and 2'3 broad. 



Occasionally spreading in 

 fruit, often erect (or suberect). 

 Petals 



On the average, about 1-0 

 mm. long and 1-5 broad.''' 

 Capsides 



On the average, about 3-5 

 mm. long, and about 1-1-1-3 

 times as long as the calyx. 



I feel that it cannot be successfully maintained that the 

 characters tabulated above are of sufficient importance to justify 

 botanists in regarding the two plants as specifically distinct. It 

 will be noticed that the differences are either comparative or refer 

 to matters of mere number and size. I find that, when large 

 numbers of specimens of the two plants are compared, the actual 

 discontinuity between the two varieties is trifling, though I admit 

 that with patient examination it is possible to determine precisely 

 almost any complete and mature specimen whether living or dried. 

 Further, if Druce is correct in identifying Kobert Brown's plant, 

 alluded to earher on in this paper, as /S. sagino'ides var. typica 

 (i.e., S. scotica Druce), then there is not merely an absence of 

 discontinuity in the important character of tlie size of the capsule, 

 but there is actually a considerable amount of overlapping. 



S. sagino'ides, in each of its British forms, may be distinguished 

 from the allied S. procumbens by its more robust habit, its larger 

 leaves which are scarcely mucronate, its usually longer pedicels, 

 its more frequently pentamerous flowers, its larger petals, its 

 larger capsules, its usually erect (or suberect) fruiting sepals, and 

 its much greater abundance in sub-alpine and truly alpine 

 localities, usually in wet situations. From the allied S. subulata 

 it may be distinguished by its being totally glabrous and eglan- 

 dular, by the much less pronounced apical mucronation of the 

 leaves, and by its smaller flowers and capsules. It has to be 

 confessed, however, that " species " in this genus have been made 

 exceedingly small, and that reduction in other parts of the genus 

 is desirable ; for example, are not S. ciliata and even S. reuteri 

 too closely allied to S. apetala to be allowed to remain as separate 

 species ? 



The two varieties of S. sagino'ides can scarcely be said to grow 

 in distinctive habitats. In Scotland, both occur in and near sub- 

 alpine and alpine streamlets and springs. The var. typica tends 

 to spread from the springs and streamsides on to the surrounding 

 siliceous grassland, and even on to small hillside Iclgers, more than 



* Petals are frequently present in S. procumbens, but are not more than 

 half as long as the sepals. 



