1 58 F. W. C. Aresclioug. 



approaches R. fruticosus, especially that form of tliis species, which grows 

 od the same tract as 11. relatus. Hut I now tliink, that R. cordifolius and 

 relatus, in regard to their origin, are, so to say, parallel forms and at the same 

 time, that R. relatus is an intermediate form hetween R. cordifolius and 

 fruticosus. As has already been mentioned, my opinion of the naturé 

 of intermediate forms, has since the printing of the first sheets of this work, 

 in some degree been modilied in consequence of the observations made last 

 summer. I am now of the opinion, that R. cordifolius has irnmi- 

 grated here from North Germany, that is to say, the form of that species, 

 which Marsson calls 11. Münteri, and under the influence of local circum- 

 stances has undergone the changes, which at present distinguish the Swedish 

 form. R. relatus, on the contraiy, I consider as a forin of R. fruticosus 

 which has arisen here, to repeat an expression, often used, a miscarried 

 atterapt of this species to produce R. cordifolius, and thus an intermediate 

 form between this species and R. fruticosus. 



Especially in Northern and Central Germany, a number of forms are to 

 be found, which belong to Cordifolii, and which on the one side approach 

 Discolores and on the other Suberecti. These forms are characterized by 

 the greater smoothness of the turiones and flowering shoots, by the sepals being 

 whito-felted on the outer side, and by the leaves being on the under side påle 

 uiicn, seldom felted. I consider that these forms deiive their origin from 

 R. fruticosus, or perhaps, even from other species of the group Suberecti. 

 On the other hand, many forms are to be found in Northern Germany 

 niiich like fchose species, but which have been developed out of, and are 

 aorthem forms of Tomentosi, which I consider has been the ca se with Disco- 

 Lores, Anmng the extraordinary number of forms which are to be found on 

 the continent, ii is, in the mean time, extremely difficult to clearly distinguish 

 between these two heterogeneous form-series. Hut in our country where these 

 groups are represented by some l<>\v forms, there is no great difficulty connec- 

 ted with it. 



Et. relä lus gives us a new proof of what 1 have öften expressed in my 

 work, thal when any new form arises from an already existing form, it nap- 

 pens al once, and noi gradually. Eowever, it must not be thought, I imagine 

 that such a sudden creation of a new form happens in the first generation. 

 ••'"i the pareni form must be imagined to have existed for a longer time on tne 

 place, where the new form is at last to arise, and during that time worked 



