432 The Systematic Value of the New Species. 



sepals, the corolla, and the calyx-tube for a certain num- 

 ber of flowers; a very great number is not necessary. 

 We plot the measurements as in the table on page 433, 

 writing opposite each length the number of flowers which 

 possess it. In the case of O. biennis I measured one 

 flower per plant; in O. muricata I did the same (I) but 

 in the case of a few plants several per plant (II). All 

 the examples of both species come from the same wild 

 locality, a sandy spot near Zandvoort (Sept. 1894). See 

 table p. 433. 



The measurements^ show in the first place that the 

 mean length of calyx and corolla for O. muricata at that 

 locality is about 14-15 mm. and for 0. biennis about 

 19-20 mm. They confirm the alleged difference between 

 the two species. But they show, further, that this differ- 

 ence is by no means such that all of the flowers of 0. 

 biennis must necessarily be larger than those of 0. muri- 

 cata or that in any given case mere size would settle the 

 question of their specific identity. On the contrary the 

 largest flowers of O. miiricata are larger than the smallest 

 flowers of O. biennis. 



The mean differences are fixed and typical. But the 

 extreme variants overlap — the ^variability is transgres- 

 sive. 



But must we conclude from this that there is no 

 boundary between the two species, that they merge into 

 one another? Not at all. For the flowers were picked 

 on undoubted muricata- and biennis--^\2ints>. 



Or we may express the case thus : the limits of the 



ica numerous other elementary species of both groups are found, but 

 they have not yet been described or named. (Note of 1908.) 



^ I have given above similar data for the fruits of Oenothera 

 leptocarpa (See page 357). 



