EP1L0BIUM DUK1/KI. 107 



quite apart from my own strong belief that E\ Duriei is entitled to 

 specific rank, it seems to me far safer to follow Gay, Barbey, and 

 others who have seen and studied the growing plant of the Pyrenees, 

 rather than any botanist, however distinguished, who bases his 

 opinion of it on herbarium study. Personally, I am almost as 

 sceptical about the existence of permanent varieties in this genus as 

 Mr. Clarke is about the occurrence of hybrids. 



Has Mr. Clarke read those parts of Haussknecht's book which 

 are written in German ? If so, I wonder at his questioning its 

 great ability and general accuracy, at any rate so far as th< 

 European part is concerned. There are some errors ; but what 

 monographer ever managed to avoid them altogether ? The more 

 I have studied our English forms, the more I have been struck 

 by Haussknecht's masterly grasp and acumen. 



With regard to the hybrid question, I must demur to Mr. 

 Clarke's description of the plan of diagnosis followed, which is 

 caricature, pure and simple. 



In some genera, hybrids are, I fully believe, rare or non- 

 existent; in some, e.g. Salix and EpUobium, I have no more doubt 

 about their frequency than about the fact that two and two make 

 four. In conversation wdth myself, Mr. Clarke has admitted that 

 they occur between the willows ; why, with no better reasons than 

 those adduced, is he so positive that there is no such thing, among 

 the willow-herbs ? In point of fact, both cases rest on the same 

 kind of proof. What Wichura did w T ith Salix, Focke has done (on 

 a smaller scale) with Epilubium ; i. e., he has produced hybrids, 

 similar to alleged natural hybrids, by artificial cross fertilisation. 



I by no means take it for granted that all seemingly inter- 

 mediate plants are bastards. These in iy be known by their total 

 (or almost total) sterility ; I understand; also, that their pollen is 

 mostly abortive, but have not proved this by microscopical experi- 

 ments. I have, however, cultivated several of these intermediates 

 for two or three seasons in succession ; among others, lan&olatum 



X obseiirum, tanceolatum x montanu**, obscurum x ptrviflorwn, and 

 purvt riorum x roumm. They have thoroughly retiiaed their peculiar 

 characters, and none of them, so far as I have noticed, has ever 

 produced perfect seeds ; whereas the parent species fruit freely, and 

 seedlings spring up all over the place. In 1890, I collected, on or 

 near Ben More of Assynt, plants which Haussknecht agreed with 

 me in naming ahiwfuUum x arwgalUdifotium, ahinefohum xpalustre, 

 and avagallidifolium x abteurum; in each case, associated with the 

 species 'in question. Specimens of all three were sent home to 

 grow, and I watched their development from time to time during 

 the following autumn, spring, and summer. In every case the 

 stolon-characters were about intermediate between those of the 

 parents ; and, although the facie* was considerably modified by 

 the changed conditions of growth, their origin was even more 

 markedly evident than before. They all remained perfectly barren ; 

 whereas E. alsinefoliunu E. anat/allitii/ulium, and E. palwstre, grown 



near them for purposes of comparison, produced normal capsules 

 and (apparently) normal seeds. 



