Compensatory Regulation. 75 



less, the morphological series is so complete that there is suffi- 

 cient ground for the conclusion that the opercula arose in the 

 course of phylogeny as modified branchiae. 



The functional operculum of Hydroides has been shown to 

 arise in ontogeny as a branchia, which later is modified to serve 

 as an operculum. Moreover, in its modification it passes through 

 a series of changes which corresponds very closely with the similar 

 morphological series that may be picked up from different genera 

 of the family. We have thus first, a stage in which the branchia 

 is unmodified (Protula stage); second, a stage with a terminal 

 enlargement on a branchial stalk having the ordinary respiratory 

 filaments still present (Filograna or Apomatus stage); third, a stage 

 in which the respiratory filaments have disappeared and in which 

 the terminal cup has only one row of serrations (Serpula stage). 

 The adult Hydroides condition, with two rows of serrations, is 

 then finally attained as the result of the reversal of the opercula 

 already described, (p. 50.) It is seen that this ontogenetic 

 series corresponds very closely with the probable phylogenetic 

 one. 



The rudimentary operculum develops as a branchia which 

 drops off and regenerates a rudimentary structure from the basal 

 stump, so that it also passes through a stage which may be con- 

 sidered to resemble a phylogenetic one. 



The regeneratory development falls under two heads, the develop- 

 ment of the rudimentary operculum from the stump of the old 

 functional or old rudimentary and the later differentiation of this 

 into a functional. The two stages, according to the conditions, 

 follow one another without a break or else there is a period of 

 rest at the rudimentary stage. The course of regeneration is 

 characterized by a great condensation and directness of the develop- 

 ment. There is no trace of a branchial stage and the development 

 of the two rows of processes of the terminal cup does not follow 

 in the ontogenetic order, that is, the more proximal row (Serpula 

 stage) does not appear before the more distal row (Hydroides 

 stage). The time between the appearance of the two rows 

 is not great but the more distal row appears before the more 

 proximal one, contrary to the course in ontogeny. The absence 

 of a branchial stage and the reversal in the order of appear- 

 ance of the two rows of processes of the cup, therefore, show 

 a wide departure from the ontogenetic development. 



