here a case of heterogenesis nearly related to the seasonal dimorphism of butterflies in a twofold manner--first, because the cycle of generations is also in this case brought about by the direct action of the external conditions of life; and secondly, the winter form is here also the primary, and the summer form the secondary one. In accordance with the idea first introduced into science by Rudolph Leuckart, we have hitherto understood heterogenesis to be only the alternation of dissimilar sexual generations. From this point of view the reproduction of _Leptodora_ can be as little ascribed to heterogenesis as can that of _Aphis_ or _Daphnia_, although the apparent agamic reproduction of the winter and a portion of the summer generation is undoubtedly parthenogenesis and not propagation by nursing.[48] As has already been said, however, I would attribute no fundamental importance to the criterion of agamic reproduction--the more especially because we are ignorant of the physiological significance of the two modes of propagation; and further, because this principle of classification is entirely external, and only valuable in so far as no better one can be substituted for it. A separation of the modes of cyclical propagation according to their genesis appears to me--especially if practicable--not alone to be of greater value, but the only correct one, and for this the knowledge of the origin of seasonal dimorphism seems to me to furnish a possible method. If, as was indicated above, we designate as metagenesis (in the narrow sense) all those cases in which it must be admitted that a series of differently aged phyletic stages have furnished the points of departure, and as heterogenesis those cases in which similar phyletic stages have been compelled to produce a cycle of generations by the periodic action of external influences, it is clear that the scope of heterogenesis is by this means considerably extended, and at the same time sharply and precisely defined.