724 



SCIENCE. 



[N. S. Vol. XV. No. 384. 



botanist has not been entangled in the 

 species of Hieracium, or who is able to 

 recognize at first sight the closely related 

 forms of Cochleariaf 



Because of the dying out of intermediate 

 forms, more ancient species may be widely 

 separated. On the other hand, more recent 

 species, whose ancestors are still alive, may 

 form narrow groups because of and with 

 these surviving ancestors. Good illustra- 

 tions of the latter are yielded by roses, 

 willows and brambles, as shown by the 

 facility with which the closely related 

 forms can be cross-fertilized, as well as by 

 the great trouble the numerous bastards 

 cause in determination. Such genera are 

 found everywhere in the plant kingdom; 

 the gentians of the Alps, for instance, or 

 the Heliantheniums, which with us seem to 

 be composed of fairly distinct types. 

 Everything indicates that in these cases the 

 species are of more recent date, and that 

 only through the dying out of intermediate 

 forms the differences between the remain- 

 ing ones have attained that degree of dis- 

 tinctness which so greatly facilitates the 

 separation of the other groups. 



In this i-egard the Oenotheras agree 

 exactly with what may be observed in 

 nature. Recent forms group themselves 

 around the mother form with minute, 

 hardly perceptible gradations. 



Once formed, the new species are as a 

 rule at once constant. No series of genera- 

 tions, no selection, no struggle for existence 

 are needed. Each time a new form has 

 made its appearance in my garden, I have 

 fertilized the flowers with their own pollen 

 and have collected and sown the seed sepa- 

 rately. The dwarf forms produce nothing 

 but dwarfs (0. nanella), the white ones 

 nothing but white ones (0. albida), the 0. 

 gig as nothing but 0. gig as, the red-nerved 

 ones nothing but corresponding specimens. 

 But a single form made an exception. This 

 was the small 0. scintillans, the seeds of 



which produced but a percentage of scin- 

 tillans plants, but here tliis inconstancy is 

 and was as much the rule as the constancy 

 of the other species. 



As an example I may cite 0. gigas. The 

 plant is as tall as 0. Lamarckiana but has 

 a more robust stem, denser foliage, a 

 broader crown of large, widely opening 

 flowers and stouter flowering-buds. The 

 fruits attain but one half the length of 

 those of plants of the mother species and 

 consequently contain fewer seeds. But the 

 individual seeds, on the other hand, are 

 rounder, fuller and heavier. This type 

 originated in my cultures of 1895 as a soli- 

 tary specimen, which at first was over- 

 looked. At that time I desired to hibernate 

 some plants, and in the latter part of the 

 autumn chose for that purpose twelve of 

 the strongest and best developed. It was 

 only in the following summer, when the 

 plants began to flower, that I noticed that 

 one plant showed differences, the impor- 

 tance of which I did not fully realize until 

 the fruits, on ripening, became much 

 shorter and stouter than ordinarily was the 

 case. It was only then that I placed the 

 raceme in a bag so as to prevent fertiliza- 

 tion with other pollen. Afterwards this 

 seed was collected separately and in the 

 spring of 1897 so-wti in a flower bed between 

 other beds sown with seeds of the normal 

 Oenothera Lamarckiana. Immediately sub- 

 sequent to germination no difference was 

 apparent, but when the third and fourth 

 leaves unfolded it suddenly became evident 

 that a new species had originated. All 

 plants. differed from their neighbors, were 

 more robust and bore broader, dai-ker 

 leaves. Though two to three hundred in 

 number, all evidently belonged to one dis- 

 tinct type. Not having, at the time, paid 

 special attention to the mother plant, I 

 was unfortunately unable to compare the 

 latter with the type at this age. But when, 

 during the summer, first the stems and 



