191 s.] PURE SPECIES OF GENOTHERA. 235 



tensive process of degeneration, form zygotes also able to survive 

 and to develop plants diverging from the parents. 



The subject of seed sterility among the Oenotheras has scarcely 

 been touched by the students of the group and yet it seems likely to 

 become a factor of prime importance in its bearings on the problems 

 of Oenothera genetics. Any worker among these plants shortly 

 becomes aware of the fact that very many of thg seed-like struc- 

 tures which he sows fail to germinate even though seed pans are 

 kept for many weeks. De Vries makes frequent reference to the 

 facts of seed sterility and the writer has in recent years recorded 

 the number of seeds sown in cultures and the number of seedlings 

 that develop. The results are most surprising and must have sig- 

 nificance although what that may be remains for the future to dis- 

 close. A line of research has opened before us that will demand 

 a special technique, for it is not enough to know merely that certain 

 proportions of the seeds germinate within the time practicable for 

 keeping seed pans under observation. 



Seed-like structures sown on the earth are obviously lost for 

 further enquiry as to the facts of their viability; a proportion of 

 seedlings appear but as for the residue, that cannot be examined. 

 The residue may contain viable seeds the germination of which is 

 delayed, or it may consist wholly of sterile structures. We must 

 develop methods that will ensure the rapid and complete germina- 

 tion of seeds in convenient receptacles such that the residue of 

 sterile structures may be left for study after the seedlings have 

 been removed and set in the earth. By such methods cultures of 

 (Enothera may be grown in which one may feel confident that all of 

 the viable seeds have germinated since by an examination of the 

 residue it may be determined whether or not the seed-like structures 

 have embryos. It is probably safe to say that no culture of (Eno- 

 thera has as yet been described in which we may feel certain that 

 the progeny of the sowing is complete. During the past winter I 

 have tested the percentage of seed fertility in some fifty species and 

 hybrids of CEnothera germinating the seeds on pads of wet filter 

 paper in Petri dishes. With this method may advantageously be 

 combined the clever practical suggestion of De Vries ('15, p. 190) 

 of forcing water into wet seeds by air pressure thereby greatly 



