332 BOTANICAL GAZETTE [October 



all the grains are alike and all the tetrads are similar to one 

 another. There are a number of hybrid derivatives of this species 

 in cultivation. Two of these, named R. Metternichi and R. Mrs. 

 H. W. Sargent, have been examined in this connection. The 

 conditions found are exactly like those presented by Epilobium in 



■ 



the stricter sense. Some of the tetrads of pollen are entirely 

 abortive, some only partially so, and in still others all the grains 

 are perfectly developed. An extremely interesting fact in the 

 present connection is the statement in standard European works 

 on the genus Epilobium that its species all hybridize naturally 

 with one another. In our specimen of Epilobium hirsutum, we 

 obviously have to do with one of these hybrid conditions. 



We are now, without further consideration of the Onagraceae, 

 which would take us beyond the scope of the present article, in a 

 position to consider advantageously the genus Oenothera, In 

 fig. 20 is shown a portion of the anther contents of Oenothera 

 Lamarckiana. Obviously the pollen is largely abortive. In 

 fact, in his Mutationstheorie, De Vries does not conceal the fact 

 that about one-third of the pollen of this species is shriveled and 

 non-functional. Not only is this true of the species, but it is like- 

 wise true of its so-called mutants, some of the weaker of these, 

 such as 0. nanella and 0. lata, having almost no good pollen. But 

 it is not in O. Lamarckiana alone that one finds imperfect pollen. 

 All of the numerous species of the genus which I have examined 

 in this connection show more or less the same state of affairs. 

 Fig. 21 illustrates the conditions in one of our commonest Oeno- 

 theras, namely O. biennis. Even with the low magnification used, 

 it is easy to see that the contents of the eight anthers shown photo- 

 graphically are far from normal. Fig. 22 shows the conditions as 

 they appear under a higher magnification. Obviously here, as 



amount 



All 



that can be said is that the genus Oenothera is in general, judging 

 from its morphological condition, in a state of high genetical 

 impurity. Fig. 23 shows a rather low magnification of several of 

 the anthers of O. Lamarckiana. O. grandiflora is illustrated under 

 the same degree of magnification in fig. 24. In this species the 

 number of abortive pollen grains present is very small, and it is 



