i8g6. THE MEANING OF METAMORPHOSIS. 399 



demonstrated by Mademoiselle de Chauvin, be arrested in its 

 development and made to re-adopt the aquatic existence and repro- 

 duce the aquatic breathing-organs of its ancestors. It is obvious, in 

 this case, that the Salamandra atra is highly tachygenic in its develop- 

 ment, and has greatly shortened up the gill-bearing larval stage 

 necessarily passed in the water by the other species of salamanders 

 that are neither so highly specialised nor so completely fitted for 

 terrestrial existence. The same is true of the tree-frog of Martinique, 

 Hylodes martini censis, which passes its tadpole stage in the ovum, and 

 is born as a true frog. Such examples as these show that the ephebic 

 (adult) highly specialised, acquired characters which fitted these 

 forms for living upon the land have become more important, and 

 encroached upon the field occupied in the ontogeny by ancestral 

 adaptive aquatic organs and tendencies, until finally these animals 

 have ceased laying their ova in the water and become able to carry 

 them within themselves until their inherited aquatic metamorphoses 

 have been passed through. 



This law has been lately explained by one of the authors of this 

 paper as due to the prepotency of acquired characters in phylogeny. 

 Acquired characters introduced late in life, and adaptive or suitable, 

 are seized upon by heredity, acting according to the law of tachy- 

 genesis, and are necessarily repeated, if they appear at all, in a 

 certain succession with relation to other characters of the ontogeny, 

 and this succession is parallel with the modifications that have taken 

 place in the evolution of the group to which the animal belongs. 

 Thus, in each succeeding member of a series, the last or later 

 acquired modifications necessarily take up space and time previously 

 occupied in ontogenic development by ancestral characters inherited 

 from more remote ancestors. Two things cannot occupy the same 

 space, and, consequently, the latest acquired and most suitable 

 characteristics press back upon and tend to obliterate those which 

 have been acquired from more remote ancestors, and which usually 

 lie between the embryo and the middle of the neanic stage — that is, in 

 the larval and earlier adolescent sub-stages. 



This is obviously a rational corollary of evolution by descent 

 with modifications, but it was not founded upon mental processes by 

 either of its authors. On the contrary, it was laboriously worked out 

 in detail in different series of animals, and carefully grounded by 

 minute observations. It does not seem to us that the earlier 

 appearance of the ephebic characters of the true frog or salamander 

 can be spoken of as " adult metamorphosis," although in these the 

 latest acquired adaptive ephebic characters have undoubtedly 

 encroached upon the period usually considered as the larval or tadpole 

 stage, leaving merely a transient recapitulation of this condition. 

 This term, however, seems more appropriate for tachygenic types 

 of this kind than for any of the Insecta. The metamorphoses of the 

 last are exactly the reverse of the process common in Amphibia, and 



