334 ON ATAVISTIC VARIATION IN OENOTHERA CRUCIATA. 



and the purple color of the nerves and the stem. The flowers are 

 as small as those of 0. muricata, but the spikes are much more 

 slender and the fruits are less broad. In all these characters the 

 European plants correspond exactly with the description given by 

 Britton and Brown. The petals of the American type are linear 

 instead of being broad and obcordate as in the allied species. 



'As yet 1 have not had an opportunity to cultivate the original 

 wild species, but I hope to be able to do so next year, as I have 

 obtained seeds from the Adirondacks through the kindness of Dr. 

 D. T. MacDougal, of New York, and Dr. B. L. Robinson, of Cam- 

 bridge, Mass^). 



Dr. MacDougal had also the kindness to ask Dr. Britton about 

 an eventual variability of the petals, but the celebrated author of 

 the Illustrated Flora informed him that Oenothera cruciata does not 

 make broad petals in America. 



This is the essential point. For in all the cultures I have as yet 

 been able to make from seeds of this species, sent to me from different 

 botanical gardens in Europe, I found the form of the petals to be 

 varying in a high degree, so as to reach, in many individuals, the 

 same outline as is presented by the petals of the allied species, 

 0. biennis, 0. muricata and others. 



For this reason, I presume that our plants are not the typical 

 0. crucia'a of Nuttall, but a variety, which perhaps has been pro- 

 duced from it in Europe. Therefore I have called my plants 0. 

 cruciata varia, merely in order to distinguish them from the pure 

 species. How this variety may have originated of course I do not 

 know. Two possibilities present themselves. The one is by mu- 

 tation, in the same way as many of our ordinary varieties of garden 

 plants have been produced, the other is by crossing, which is perhaps 

 a still more common source of new garden varieties. 



Both possibilities seem to me to be of some interest, since they 

 bear directly on the great question of the internal causes of incon- 

 stancy in general. For in my variety the petals do not vary accor- 

 ding to Quetelet's law, about a mean, which lies somewhere between 

 the obcordate and the linear form, producing petals of which the 

 majority do not essentially differ from this mean, whilst the extremes 

 are very rare. Quite on the contrary, the obcordate and the linear 



i) Seeds from different localities would be always verj- welcome to me, 

 as the plant may be in a mutable state in some districts, while it is not 

 so in others. 



