No. 459.] ı ANATOMY OF CATALPA HYBRIDS. 133 
The question still remains unanswered, as to how far this 
hybrid approaches and satisfies the conditions which define a 
species. A complete solution of this question would be obtained 
in the most satisfactory manner by an anatomical study of 
reverted forms of the hybrid, were it possible to satisfy all the 
requisite conditions of authenticity, and determine to what 
extent the anatomical changes already found to characterize the 
hybrid, revert to the structure of the separate parents. Unfor- 
tunately such studies have not been possible under the conditions 
of the present investigation, and we are therefore compelled to 
fall back upon such partial evidence as is afforded by the exter- 
nal morphology. Reference to the original description of the 
hybrid shows that seedlings exhibit a decided tendency to 
reversion with respect to the character of the flower, and that 
such tendency is always in the direction of C. kempferi. This 
fact makes it clear that the Japanese parent exerts a dominant 
influence, while it also shows that the characters are not always 
fully fixed. Reference to the original description, however, 
makes it clear that such reversion is not displayed by aX seed- 
lings but only by some, results which are within the limits of 
justifiable anticipation and in strict accord with Mendel’s law. 
Possibly our knowledge of all the changes incident to reversion 
is not as yet sufficiently complete to admit of a final expression 
of opinion as to the actual stability of the hybrid and its status 
as a species, but it may be kept in mind that while some seeds 
exhibit reversion, others come true to the hybrid form which is 
capable of perpetuating itself in succeeding generations. There 
is in this case no reason for supposing that this is any less a true 
species than large numbers of plants in which hybridization is 
an important factor as the starting point for a new line of devel- 
opment. With the abandonment of the old idea of fixity of 
species, and more particularly in the light of Mendel’s law, we 
are led to see that in any case of hybridization there must be 
large numbers of progeny which, through reversion, ultimately 
disappear, while comparatively few acquire such stability as to 
survive in the form of what must be regarded as definite species. 
It is only on such grounds that we are justified in regarding as 
species the numerous and but slightly differentiated forms of the 
