490 



SCIENCE 



[N. S, Vol. XXXIX. No. 1005 



gation in cultures. Professor Brainerd has 

 been kind enough to supply some flowers of 

 these interesting specimens for the present 

 work and the pollen conditions are clearly 

 those found in hybrids. 



We may now profitably turn our attention to 

 the Onagraceae, to which the well-advertised 

 (Enothera lamarcTciana belongs. Our common 

 garden fuchsias are known to be of hybrid 

 origin. In some of the varieties the pollen 

 sterility is almost complete. This is notably 

 the case in a hybrid derivative of Fuchsia 

 triphylla, grown in the Harvard garden. In 

 other varieties the sterility is often less marked. 

 This condition has been found to obtain for 

 example in long-tubed fuchsia hybrids. Here 

 a certain number of the grains are perfectly 

 developed and have normal protoplasmic con- 

 tents, while others are small, shrivelled, col- 

 lapsed and without protoplasmic contents. 

 Among the wild-growing species of the Ona- 

 graceae, those of Epiloiium have long been 

 recognized by European systematists as prone 

 to produce spontaneous hybrids. A good illus- 

 tration in the present connection is supplied 

 by a hybrid form of E. hirsutum, occurring 

 commonly in the vicinity of ballast heaps. 

 Here the pollen is to a large extent abortive, 

 the degeneracy sometimes affecting the whole 

 of the contents of the anther sack or in other 

 cases being confined to a greater or smaller 

 number of the grains. In contrast to E. hir- 

 sutum may be mentioned our common fire- 

 weed, E. angustifolium (sometimes put under 

 a different genus). In all the abundant mate- 

 rial of this species examined the pollen was 

 entirely normal. Indications of hybridization 

 correlated with corresponding pollen condi- 

 tions have also been noted in the case of other 

 representatives of the Onagraceae, but the 

 illustrations mentioned will sufiice for our 

 present purposes. 



We may now turn to the genus (Enothera 

 itself. In his " Mutationstheorie " De Vries 

 has noted that about one third of the pollen 

 of 0. lamarcTciana is sterile, and abortive. 

 This statement I can only confirm. Even in 

 the more vigorous of the so-called mutants 

 originating in cultures of 0. lamarckiana from 



the seed, the pollen is very largely degenerate 

 and in the less vigorous elementary species 

 often almost completely so. In 0. lata (mu- 

 tant of 0. LamarcJciana) the pollen is fre- 

 quently entirely sterile. But it is not only in 

 0. lamarchiana and its so-called mutants or 

 elementary species, that pollen sterility is to be 

 seen, for this condition is well nigh universal 

 in the species of this Onagraceous genus, 

 recognized in systematic works. For example 

 in the very common and variable species 

 known as 0. biennis, half of the pollen grains 

 are sometimes abortive. This condition I have 

 observed in specimens from regions as far 

 apart geographically as the state of Massachu- 

 setts, the province of Ontario and the shores 

 of the Gulf of St. Lawrence. The examina- 

 tion of a large amount of material of recog- 

 nized wild species of Oenothera has led me to 

 the apparently inevitable conclusion that spon- 

 taneous hybridism is extremely common in the 

 genus and that in general it represents a con- 

 dition of high genetical impurity. The purest 

 species which has come under my notice, so far 

 as may be judged from the pollen conditions 

 is 0. grandiflora, obtained from Tensaw. Here 

 pollen abortion is well nigh absent both in 

 flowers gathered for me by Professor Tracey 

 and in those produced from seed both by Pro- 

 fessor Davis and myself. Continued growth 

 in the Harvard botanic garden has not altered 

 this characteristic in any degree, for speci- 

 mens grown last summer show the same con- 

 dition of relative genetical purity or at least 

 freedom from inharmonious hybridization. 



It may be argued by some that the more or 

 less marked constancy of the generally ac- 

 cepted species of (Enothera makes it clear that 

 they are normal species. It is now recog- 

 nized however that constant hybrids are of 

 extremely common occurrence both in nature 

 and as a result of experimental crossing. This 

 is particularly true of species crosses. There 

 is consequently no good reason why we should 

 not admit that the genus (Enothera is strik- 

 ingly characterized by spontaneous hybridiza- 

 tion. There appears in fact to be every rea- 

 son to believe that the bar sinister has been 

 crossed and double crossed in our American 



