XIV PROCEEDINGS OE THE 



and a species ; and tlie exceptional cases he has observed, more 

 or less interfering with any universal regularity in the rules above 

 given, are some of them very striking. 



One new point brought out by the observations of both these 

 experimenters is, that when hybrids between species or marked 

 varieties do become fertile, that fertility increases in successive 

 generations, accompanied, apparently, by a tendency more or less 

 rapid to return to one of the parent types. M. G-odron, reasoning 

 a priori from the fact, supposed to be proved, that hybrids are 

 never fertile except when fecundated by the pollen of one of the 

 parents, regards this return to the parent type and increasing 

 fertility as the natural consequence of such fecundation ; but 

 Naudin states that this increased fertility and return occur 

 constantly (though not uniformly) under conditions in which he 

 believes that no such fecundation from the parent type can 

 have occurred. This he is disposed to account for by the supposi- 

 tion that the two parent essences are intermixed, not truly com- 

 bined, in the hybrid and in all its parts, reproductive as well 

 as vegetative, and that, in future generations, the tendency 

 of the reproductive parts of one essence to combine with those 

 of the same essence has the gradual effect of eliminating the 

 other. This hypothesis appears to be new, and many circum- 

 stances are adduced in its support, but it requires much farther 

 observation and a much better knowledge than we yet possess 

 of the physiology of fecundation before it can be admitted as 

 proved. 



Puzzling exceptional cases occur to both experimenters, in 

 which, as usual, if they cannot refute the conclusions to be drawn, 

 they are obliged to doubt the premises. Thus ^gilops speltce- 

 formis, asserted to be the readily bred offspring of the hybrid ^. 

 triticoides fecundated by wheat, its male parent-type, is found 

 both by Naudin and G-odron to be uniformly fertile without fur- 

 ther assistance from either parent, and to remain unchanged 

 through successive generations. G-odron, who admits that hy- 

 brids fecundated by one of the parents produce offspring which 

 have so far returned to the nature of species as to become fertile, 

 cites this in support of his rule that there is no return to the na^ 

 ture of the original parent without fecundation by that parent. 

 Naudin doubts the assumed origin of ^. speltceformis. So again, 

 Naudin, among the observations in support of his theory of inter- 

 mixture of species, adduces the fact, not only of the well-known 

 differences in resemblance to either parent in their hybrid off- 



