November 12, 1909] 



SCIENCE 



693 



were alwaj'S dry and empty) from several 

 flowers of an individual of 0. lata, at the 

 same time removing the stigma and style by 

 pulling the latter out at the base as an extra 

 precaution, afterwards covering the flower 

 with a bag and marking the capsule according 

 to the method I ordinarily use in making 

 guarded crosses. All of the flowers so treated 

 but one gave negative results, but this one 

 produced three fair-sized seeds. 



Ordinarily, if, for some reason, a flower fails 

 to be pollinated, the ovules remain, very small 

 and gradually dry up and wither, so that after 

 a few weeks such an ovary has not grown in 

 size and if broken open shows numerous small, 

 dried granules which are the remnants of the 

 deteriorated ovules, many of them still at- 

 tached in their original position. These three 

 seeds, while slightly below the average in size, 

 yet were hundreds of times larger than the 

 small remnants of such unfertilized ovules, 

 and indeed there were many of the latter in 

 the capsule in question, in addition to the 

 three seeds. 



In every case where pollination was thus 

 prevented, the ovary remained very small and 

 gradually dried up and shrank to a small 

 diameter, and the one containing the seeds 

 was but little larger than the rest. Several of 

 these small dry ovaries fell off and hence were 

 never examined for seeds. The number of 

 seeds, if there were any present, could not 

 have been large in any of them. 



I also treated, in a similar manner, a num- 

 ber of flowers from several individuals of the 

 English 0. lata, which produces some pollen; 

 but without exception the results were nega- 

 tive. 



In this connection wiU be recalled the dis- 

 covery of Ostenfeld' and Eosenberg' that cer- 

 tain species of Hieracium are partly apoga- 

 mous or aposporous, and partly require fer- 

 tilization. But in this genus of Composites, 

 where each flower of a head develops a single 

 seed which is independent of all the other seeds 



' " Castration and Hybridization Experiments 

 with some Species of Uieracia," Bot. Tidsskrift, 

 27: 225-248, 1906. 



• " Cytological Studies on the Apogamy in Hier- 

 acium," Bot. Tidsskrift, 28: 143-170, 1907. 



of a head, the conditions of nutrition are much 

 more favorable to partial apogamy when pollen 

 is excluded from the head, than is the case in 

 an Oenothera capsule where the ovules are 

 closely crowded together into four chambers 

 and the deterioration of the great majority of 

 them in the absence of fertilization is likely 

 to carry down the others in the common ruin 

 and also to lead to the cutting oif of the com- 

 mon food supply. 



So far as I am aware, the only other indi- 

 cation of the development of embryos in 

 (Enothera without previous fertilization is in 

 0. gigas. Schouten' reports obtaining one 0. 

 Iwvifolia individual in a large culture of 0. 

 gigas. Now I have found that 0. laivifolia 

 has fourteen chromosomes, while 0. gigas is 

 known to have twenty-eight." Such an indi- 

 vidual of 0. Iwvifolia might have arisen from 

 0. gigas through a process of parthenogenesis 

 in the restricted sense of Strasburger,' an egg 

 with the reduced number of chromosomes 

 producing the embryo without fertilization. 

 At present no case of this sort is known in 

 the plant kingdom, although in echinoderms 

 and various other animals the artificial pro- 

 duction of larvse from unfertilized eggs is a 

 well-knovm fact and, in some of these cases at 

 least, the number of chromosomes is the re- 

 duced number. Whether the origin of this 0. 

 losvifolia individual was of a similar sort must 

 remain for the present undecided. The fact 

 that in such plant genera as Alchemilla and 

 Hieracium the apogamous members of the 

 genus frequently have about twice as many 

 chromosomes as the normally fertilized mem- 

 bers would make the occurrence of similar 

 conditions in 0. gigas a thing which might 

 reasonably be anticipated. 



This indication of apogamy in 0. lata of 



' " Mutabiliteit en variabiliteit," p. 93, disserta- 

 tion, Groningen, 1908. 



' In all these forms there are probably occasional 

 departures of one or more chromosomes from the 

 usual number, owing to the occasional irregulari- 

 ties in chromosome distribution which I have 

 shown {Bot. Oaz., 46: 1-34) to occur in the forma- 

 tion of the germ cells. 



' " Apogamie bei Marsilia," Flora, 97 : 163, 

 1907. 



