80 THE JOURNAL OF BOTANY 



essential to remove the confusion and vague speculation which 

 are so rife in this part of the subject, and which have assisted 

 perhaps more than anything else in bringing the study of plants 

 in the field into disrepute among specialists in other branches 

 of botany. 



E. adnatum Griseb. f. stenophylla Hausskn. $ xE. hirsutum 



Linn. 2 • 

 This hybrid (I. c. p. 161) keeps quite constant as regards 

 essential features, but shows a striking amount of variation in 

 vegetative characters, especially among the young spring shoots. 

 Some of these are much more slender than the main stems, and 

 the leaves are narrower, relatively longer, and more deeply 

 toothed : these characters are not permanent, however, and 

 cannot be regarded as indicating bud- sports. 



E. hirsutum Linn. ? x E. montanum Linn. $ . 



This cross was made in 1909, and the two plants which 

 resulted have been propagated by division and kept under 

 observation during three summers, 1910-1912. Dried specimens 

 of this and the two following artificially produced hybrids are 

 being distributed through the Botanical Exchange Club. This 

 is perhaps the most remarkable of the Epilobium hybrids which 

 I have grown ; it is quite unlike what would be prophesied to 

 occur from its parentage if the method of intermediates were 

 adopted ; and in fact the herbarium specimens and exsiccata 

 which I have seen labelled E. hirsutum x E. montanum are 

 quite distinct from the experimentally produced hybrid. Hauss- 

 knecht * identifies the plant known as E. purpureum Fries as 

 this hybrid, rejecting the opinions of earlier authorities ; viz., 

 Fries, who had named E. purpureum a plant which Hauss- 

 knecht considers to be E. palustre x E. roseum, in addition to 

 another plant identified as E. hirsutum x E. montanum; and 

 Nyman, who thought E. purpureum was possibly E. montanum x 

 E. tetragonum. The description given by Haussknecht agrees 

 fairly well, so far as it goes, with my artificially produced 

 E. hirsutum $ x E. montanum $ , and it seems probable that 

 his identification of E. purpureum is correct, though there are 

 certain discrepancies. I have not seen specimens of E. purpureum 

 Fries, but I have no hesitation in saying that the large-flowered 

 plants with ovate woolly leaves labelled E. hirsutum x E. mon- 

 tanum, which I have seen in English herbaria, are wrongly 

 identified. 



My two plants during their first year (1910) showed a remark- 

 able feature which has been repeated in 1911 and 1912 ; this was 

 the production of two kinds of shoots as follows : — 



(i) The first stems to appear from the rosettes are slender and 

 twiggy, bearing comparatively closely set small lanceolate leaves, 

 4-5 cm. in length x 1 cm. broad ; these stems branch repeatedly, 



* Monographic der (rattling Epilobium, Jena, 1884, p. 63. 



