ON THE VARIATIONS OF THE EVENING PRIMROSE 355 



In the praiseworthy attempt to settle the question of the 

 original habitats of the various forms of Onagras which have been 

 brought to the front by De Vries's work on mutation, he has 

 given a detailed description of what lie regards as " CE. biennis (in 

 the strictest sense) " from plants growing on uncultivated land in 

 the New York Botanical Garden. His description of the plant, 

 stated to be hirsute, with finely pubescent buds and small 

 flowers with petals only 12 to 16 millim. long, and sepals much 

 shorter than the tubular portion of the hypanthium, with the pistil 

 constantly shorter than the stamens, did not agree well with the 

 Evening Primrose I had under observation. I therefore submitted 

 a plant to Prof. De Vries, who kindly informed me that my Evening 

 Primrose is a " CE. biennis L. as it occurs everywhere in Europe; 

 it is probably the type described by Linn6. The CE. biennis of 

 Dr. MacDougal is another elementary species, occurring in the 

 United States of America. It is one of the numerous forms of 

 CE. biennis found in the United States, but which has not yet 

 been studied or described. , ' Two flowers from the Botanic Garden 

 of Amsterdam, which Prof. De Vries was so good as to send me 

 for comparison, differ only in being a little larger, the petals 

 measuring 27 millim. against 24, which is the maximum length in 

 the flowers observed by me. It is therefore with the sanction of 

 this great authority that I designate the CEnothera growing in the 

 Natural History Museum grounds as the true CE. biennis. 



I have found it impossible to distinguish any well-marked 

 varieties among the plants which grow under my eyes. And yet 

 they present a great amount of variation, which I think deserves 

 being placed on record. 



In some plants the buds are closely grouped together, in others 

 they stand wide apart, forming or not long spikes above the open 

 flowers ; in some the stalks or the midrib of the leaves are pink 

 or crimson, in others there is no trace of such colour. But it is 

 especially in the flowers that the variations are most striking if 

 they be compared with De Vries's account of the " elementary 

 species " with which he has dealt. The length of the petals varies 

 between 10 and 24 millimetres. The pistil may be as long as, longer, 

 or shorter than the stamens, and in some buds just before open- 

 ing it projects most distinctly. I here transcribe De Vries's 

 own words (i. p. 319) in defining his CE. Lamarckiana from 

 CE. biennis : — 



"Dass sie [the new species of the Lamarckiana group] naher 

 verwandt Bind, als mit GE. biennis, CE. muricata, CE. suaveolens 

 und den sonstigen in systematischen Werken beschriebenen Arten 

 dieser Gruppe, ergiebt sich, ausser aus ihrer Abstammung, eigent- 

 lich nur aus den Eigenschaften der Bluthe. Diese sind erstens 

 viel grosser als bei den verwandten Formen, und zweitens haben 

 sie langere Griff el* Der Griff el erhebt die Narben bereits in den 



* Further on, however, one of the mutants of (E. Lamarckiana is described 

 as (E. brevistylis (ii. p. 430), in which the style does not project, or but slightly 

 projects, beyond the tube. 



2d2 



