218 THE JOURNAL OF BOTANY 



course is, however, more admissible than to call the plant E. Rost- 

 koviana. I should probably consider it a form of E. Rostkoviana, 

 and naturally leave the name-giving and publication to the dis- 

 coverer. Should it be found that the plant is widespread, and 

 takes in its distribution the place of E. Rostkoviana, its description 

 as a species would be quite as admissible as that of E. cam- 

 pestris Jord. 



" From typical E. Rostkoviana the new plant differs by its 

 smaller flowers (at the longest 8 mm.), the darker shade of its 

 corolla, its smaller leaves, and the slenderness and delicacy of all 

 its parts. Some of these marks appear in other varieties of E. 

 Rostkoviana, i. e. the small flowers in var. minorifiora Borb., as well 

 as on some Alpine species, and the purple corolla in var. rubra 

 (Baumg.) ; but, as I have said before, I know of no plant which 

 combines these characters as the one under notice. 



11 There is another reason which makes the plant more interest- 

 ing still. E. Rostkoviana is well-defined and constant in the centre 

 of the district in which it is found, but it varies very much on the 

 boundaries of that district. On the south-west border of the area 

 E. campestris has formed itself; on the east and south-east border 

 there is very often a kind which is very sparingly hairy. In the 

 north-east of the area E. fennica Kihlm. has formed itself out of 

 E. Rostkoviana. It might be possible that in the north-west of the 

 area of E. Rostkoviana such a representative plant as that now 

 before me has been formed. This view is supported by the fact 

 that I have several times seen extremely small-flowered E. Rost- 

 koviana from England. 



u Concerning the two specimens I have been dealing with, I 

 believe them to be one and the same form, the absence or presence 

 of gland-tipped hairs, as well as the difference in the time of 

 flowering [June in one case, September-October in the other] , 

 being associated with certain local conditions." 



It is deserving of record that, quite independently of Prof, von 

 Wettstein's investigations, the Rev. E. S. Marshall recently arrived 

 at the same conclusion. Writing me shortly after I had heard 

 from the Vienna botanist, Mr. Marshall said : — " I have now your 

 Porth Towan sheet, June 21st, 1905, and Dr. Vigurs's Roborough 

 sheet, September 18th, 1906, side by side before me. Allowing for 

 the difference of the season (and most of our plants flower from 

 June to September), there does not seem to be any very material 

 difference, except that the S. Devon plant has even the lowest 

 leaves more or less glandular, and, indeed, the glands are more 

 numerous throughout. . . . Nor, in the specimens before me, is 

 there any material difference in the leaf-outline ; and both have 

 very crowded leaves. • . . I have always thought that these 

 Devon and Cornwall plants came nearest to Rostkoviana of our 

 forms, but were distinct from it and brevipila. Usually, at any 

 rate in the South of England, Rostkoviana is a very well-marked 

 and fairly uniform plant, though the size of the flowers varies 

 (perhaps according to situation and luxuriance). As to the colour 

 of flowers in brevipila, I have seen a great deal of that species, and 





