LIFE-HISTORY OF THE ROCK BARNACLE. 3 



for example, may be quite different from I), and E 2 from E. When 

 modification in the development has proceeded so far, we might 

 represent the new form by such a symbol asABCPQR G. the three 

 earlier stages being supposed in this case to have remained unaltered, 

 whilst the stages T), E, and F have become so much modified that 

 new symbols P, Q, and R are required to indicate them ; where the 

 adults have changed little w r e might get such formulae as A Q C D, 

 PQRD, &c. In such cases the development will only to a very 

 limited extent express the history of the race. This " falsification " 

 of the record was recognised many years ago, and Haeckel showed 

 that in considering the development of any animal we must dis- 

 tinguish characters which were palingenetic, or ancestral, and 

 characters which were cenogenetic, or acquired. 



A further set of disturbances may be, and very commonly actually 

 appear to be, set up in the development by the early or precocious 

 appearance in the young form of characters belonging to the adult. 

 Thus we may get sequences such as the following : — A B CD, 

 Ai Bi d D ls and A x di d 2 D 2 , the larval forms B and C being replaced 

 by di and d 2> which already show many characters originally confined 

 to the later stage D. 



The problem as to the value of a given larval form is thus a very 

 complicated one, and we have to face the question as to how we can 

 eliminate the cenogenetic or acquired larval characters. In the case 

 of relatively rapidly acquired characters this can sometimes be done 

 by comparing the development with that of remaining members of 

 the group : if the embryos and larvae of a form show characters which 

 differ widely from those seen in the young forms of the rest of the 

 group, we. may reasonably believe that the development has been 

 modified, especially if we can point to special features in the environ- 

 ment which would be likely to bring about such changes, or if we have 

 reason to believe from the structure of the adult that w ? e are not 

 dealing with one of the most primitive forms of the group. Much 

 may be learnt of the primitive mode of development by seeing which 

 characters are common to the embryos and larvae of the various mem- 

 bers of the group, and valuable clues may be obtained by studying 

 the development of the form to which adult anatomical structure 

 points as the most primitive ; for supposing all stages to vary equally 

 in a given length of time, we may reasonably conclude that the 

 adults which have been least modified will have the least modified 

 larva?. 



Let us suppose that we have succeeded in eliminating many of 

 the cenogenetic characters from the development, say, of the Barnacle, 

 and have an idea of the series of young forms passed through by the 



